en MOLSIM is a domain (application) ontology designed to semantically represent platform-agnostic biomolecular simulations (all-atom and coarse-grained) as FAIR (Findable, Accessible, Interoperable, and Reusable) datasets. As an application ontology, it reuses terms from established OBO ontologies where they fit and defines its own where they do not, aiming to standardize the representation of biomolecular simulation data, processes, and methodologies across platforms and tools. biomolecular simulation Molecular Simulation Ontology molsim AI assistance disclosure: initial drafts of the term definitions in this ontology were produced with the assistance of large language models, primarily Google Gemini, with further drafting and curation support from Anthropic Claude; all drafts were then curated and reviewed by humans. Verification status is recorded per term and is machine-readable: a term carrying an IAO:0000117 (term editor) annotation has been checked by the named domain expert, whose ORCID is given; a definition whose text still ends with the marker (UNVERIFIED) has not yet been reviewed by a domain expert and should be treated as a draft. Terms marked owl:deprecated are retired and are exempt from verification. Release-format note: the OWL and JSON products (molsim.owl, molsim.json and their -base/-full variants) are the complete and canonical releases. The OBO-format product (molsim.obo) is a CLASS-ONLY VIEW: OBO 1.2 cannot represent datatype properties, and named individuals are not carried through the conversion, so the data properties (per-species composition counts, boolean flags, serialised arrays) and the named individuals of this ontology are absent from it. The class hierarchy, definitions, synonyms and object properties in molsim.obo are complete and correct. Use the OWL or JSON release if you need the full content. Scope note: MOLSIM's boundary is drawn by USE CASE, not by subject matter. The membership test for a new term is: would a complete record of a simulation dataset need to state it? That spans what was configured (timestep, thermostat, force field), what was computed (RMSD, potential of mean force, free energies), where the inputs came from (a PDB entry, a UniProt accession, a publication), and what it ran on (engine, version, hardware). Terms from neighbouring fields (quantum chemistry, computational materials science, structural biology) are admitted at most two levels below their attachment point: we record which QM engine ran and which basis set parameterised a force field, but not the internal machinery of quantum-chemical methods. This rule governs terms added from 2026-07-29 onward; earlier terms are retained. MOLSIM is the OBO-aligned ontology for biomolecular simulation metadata and is complementary to EMMO-based materials-modelling ontologies (EMMO, OSMO, VISO), with which it interoperates by SSSOM mapping rather than by adopting a shared upper ontology. Full statement: see the README. 2026-10-16 https://github.com/CPCLab/molsim-ontology editor preferred label editor preferred label editor preferred term editor preferred term editor preferred term~editor preferred label The concise, meaningful, and human-friendly name for a class or property preferred by the ontology developers. (US-English) PERSON:Daniel Schober GROUP:OBI:<http://purl.obolibrary.org/obo/obi> editor preferred term example example of usage A phrase describing how a term should be used and/or a citation to a work which uses it. May also include other kinds of examples that facilitate immediate understanding, such as widely know prototypes or instances of a class, or cases where a relation is said to hold. PERSON:Daniel Schober GROUP:OBI:<http://purl.obolibrary.org/obo/obi> example of usage has curation status PERSON:Alan Ruttenberg PERSON:Bill Bug PERSON:Melanie Courtot OBI_0000281 has curation status definition textual definition The official definition, explaining the meaning of a class or property. Shall be Aristotelian, formalized and normalized. Can be augmented with colloquial definitions. 2012-04-05: Barry Smith The official OBI definition, explaining the meaning of a class or property: 'Shall be Aristotelian, formalized and normalized. Can be augmented with colloquial definitions' is terrible. Can you fix to something like: A statement of necessary and sufficient conditions explaining the meaning of an expression referring to a class or property. Alan Ruttenberg Your proposed definition is a reasonable candidate, except that it is very common that necessary and sufficient conditions are not given. Mostly they are necessary, occasionally they are necessary and sufficient or just sufficient. Often they use terms that are not themselves defined and so they effectively can't be evaluated by those criteria. On the specifics of the proposed definition: We don't have definitions of 'meaning' or 'expression' or 'property'. For 'reference' in the intended sense I think we use the term 'denotation'. For 'expression', I think we you mean symbol, or identifier. For 'meaning' it differs for class and property. For class we want documentation that let's the intended reader determine whether an entity is instance of the class, or not. For property we want documentation that let's the intended reader determine, given a pair of potential relata, whether the assertion that the relation holds is true. The 'intended reader' part suggests that we also specify who, we expect, would be able to understand the definition, and also generalizes over human and computer reader to include textual and logical definition. Personally, I am more comfortable weakening definition to documentation, with instructions as to what is desirable. We also have the outstanding issue of how to aim different definitions to different audiences. A clinical audience reading chebi wants a different sort of definition documentation/definition from a chemistry trained audience, and similarly there is a need for a definition that is adequate for an ontologist to work with. 2012-04-05: Barry Smith The official OBI definition, explaining the meaning of a class or property: 'Shall be Aristotelian, formalized and normalized. Can be augmented with colloquial definitions' is terrible. Can you fix to something like: A statement of necessary and sufficient conditions explaining the meaning of an expression referring to a class or property. Alan Ruttenberg Your proposed definition is a reasonable candidate, except that it is very common that necessary and sufficient conditions are not given. Mostly they are necessary, occasionally they are necessary and sufficient or just sufficient. Often they use terms that are not themselves defined and so they effectively can't be evaluated by those criteria. On the specifics of the proposed definition: We don't have definitions of 'meaning' or 'expression' or 'property'. For 'reference' in the intended sense I think we use the term 'denotation'. For 'expression', I think we you mean symbol, or identifier. For 'meaning' it differs for class and property. For class we want documentation that let's the intended reader determine whether an entity is instance of the class, or not. For property we want documentation that let's the intended reader determine, given a pair of potential relata, whether the assertion that the relation holds is true. The 'intended reader' part suggests that we also specify who, we expect, would be able to understand the definition, and also generalizes over human and computer reader to include textual and logical definition. Personally, I am more comfortable weakening definition to documentation, with instructions as to what is desirable. We also have the outstanding issue of how to aim different definitions to different audiences. A clinical audience reading chebi wants a different sort of definition documentation/definition from a chemistry trained audience, and similarly there is a need for a definition that is adequate for an ontologist to work with. PERSON:Daniel Schober GROUP:OBI:<http://purl.obolibrary.org/obo/obi> definition definition editor note An administrative note intended for its editor. It may not be included in the publication version of the ontology, so it should contain nothing necessary for end users to understand the ontology. PERSON:Daniel Schober GROUP:OBI:<http://purl.obfoundry.org/obo/obi> GROUP:OBI:<http://purl.obofoundry.org/obo/obi> editor note term editor Name of editor entering the term in the file. The term editor is a point of contact for information regarding the term. The term editor may be, but is not always, the author of the definition, which may have been worked upon by several people 20110707, MC: label update to term editor and definition modified accordingly. See https://github.com/information-artifact-ontology/IAO/issues/115. PERSON:Daniel Schober GROUP:OBI:<http://purl.obolibrary.org/obo/obi> term editor alternative term An alternative name for a class or property which means the same thing as the preferred name (semantically equivalent) PERSON:Daniel Schober GROUP:OBI:<http://purl.obolibrary.org/obo/obi> alternative term definition source Formal citation, e.g. identifier in external database to indicate / attribute source(s) for the definition. Free text indicate / attribute source(s) for the definition. EXAMPLE: Author Name, URI, MeSH Term C04, PUBMED ID, Wiki uri on 31.01.2007 PERSON:Daniel Schober Discussion on obo-discuss mailing-list, see http://bit.ly/hgm99w Discussion on obo-discuss mailing-list, see http://bit.ly/hgm99w GROUP:OBI:<http://purl.obolibrary.org/obo/obi> definition source term replaced by Use on obsolete terms, relating the term to another term that can be used as a substitute Person:Alan Ruttenberg Person:Alan Ruttenberg Add as annotation triples in the granting ontology term replaced by A relation that gives the name a simulation program uses for the concept the annotated term stands for, as that name appears in the program's input file, output file or command line. The program the name belongs to is recorded as the source of each such statement, so 1 property serves every program. (UNVERIFIED) has software keyword an alternative term used for STATO statistical ontology and ISA team Alejandra Gonzalez-Beltran Orlaith Burke Philippe Rocca-Serra STATO alternative term a R command syntax or link to a R documentation in support of Statistical Ontology Classes or Data Transformations Alejandra Gonzalez-Beltran Orlaith Burke Philippe Rocca-Serra R command an annotation property to provide a canonical command to invoke a method implementation using Python programming language Alejandra Gonzalez-Beltran Orlaith Burke Philippe Rocca-Serra Python command created by creation date has_alternative_id has_broad_synonym database_cross_reference An alternative label for a class or property which has the exact same meaning than the preferred name/primary label. has exact synonym has_exact_synonym has_narrow_synonym has_obo_namespace has_related_synonym id in_subset label label a core relation that holds between a whole and its part Everything has itself as a part. Any part of any part of a thing is itself part of that thing. Two distinct things cannot have each other as a part. Occurrents are not subject to change and so parthood between occurrents holds for all the times that the part exists. Many continuants are subject to change, so parthood between continuants will only hold at certain times, but this is difficult to specify in OWL. See http://purl.obolibrary.org/obo/ro/docs/temporal-semantics/ Parthood requires the part and the whole to have compatible classes: only an occurrent have an occurrent as part; only a process can have a process as part; only a continuant can have a continuant as part; only an independent continuant can have an independent continuant as part; only a specifically dependent continuant can have a specifically dependent continuant as part; only a generically dependent continuant can have a generically dependent continuant as part. (This list is not exhaustive.) A continuant cannot have an occurrent as part: use 'participates in'. An occurrent cannot have a continuant as part: use 'has participant'. An immaterial entity cannot have a material entity as part: use 'location of'. An independent continuant cannot have a specifically dependent continuant as part: use 'bearer of'. A specifically dependent continuant cannot have an independent continuant as part: use 'inheres in'. has part x precedes y if and only if the time point at which x ends is before or equivalent to the time point at which y starts. Formally: x precedes y iff ω(x) <= α(y), where α is a function that maps a process to a start point, and ω is a function that maps a process to an end point. precedes has measurement unit label This document is about information artifacts and their representations A (currently) primitive relation that relates an information artifact to an entity. 7/6/2009 Alan Ruttenberg. Following discussion with Jonathan Rees, and introduction of "mentions" relation. Weaken the is_about relationship to be primitive. We will try to build it back up by elaborating the various subproperties that are more precisely defined. Some currently missing phenomena that should be considered "about" are predications - "The only person who knows the answer is sitting beside me" , Allegory, Satire, and other literary forms that can be topical without explicitly mentioning the topic. person:Alan Ruttenberg Smith, Ceusters, Ruttenberg, 2000 years of philosophy is about A relation that links a molecular simulation system to the geometric shape of the simulation cell used to apply periodic boundary conditions. Those conditions mimic an infinite system by making a particle that leaves one side re-enter from the opposite side. Common shapes are cubic, triclinic and rhombic dodecahedron, and the choice affects both computational cost and how isotropic the system is. (UNVERIFIED) has periodic box type A relation that links a computational process or analysis step to the utility program from a molecular dynamics software suite that was used to perform it. For example, a trajectory analysis may use a particular trajectory processing utility, or system preparation a particular solvation tool. It records the exact tool used within a larger workflow, where that tool is one component of a bigger software package. (UNVERIFIED) has molecular dynamics suite utility A relation that links a structural alignment procedure or analysis step to the software program used to perform the alignment. Structural alignment superimposes molecular structures such as proteins or nucleic acids, so that their conformations can be compared or conserved regions found. (UNVERIFIED) has alignment tool A relation that links a molecular structure to the origin from which its three-dimensional coordinates were obtained. The origin may be an entry in an experimental database such as the Protein Data Bank, the output of a computational prediction method such as homology modelling, or another defined source. Recording it lets a reader trace where a structure came from and judge its quality. (UNVERIFIED) has protein structure source A relation that links a molecular simulation system to the overall physical context or medium it represents, such as an aqueous solution, a lipid bilayer, a vacuum or a crystal lattice. That context shapes the choice of simulation parameters, including boundary conditions and dielectric treatment, and how the results should be read. has simulation environment A relation that links a molecular simulation system to the initial three-dimensional structural model of the molecular complex used as its starting point. That structure is often taken from experimental data such as a PDB entry, or from earlier computational modelling such as a docking pose. (UNVERIFIED) has template complex structure A relation that links a plan specification to the rules, checks or conditions applied to confirm that the chosen ligands, and the alchemical transformations between them, are suitable for a relative binding free energy calculation. Those checks may cover structural similarity, consistency of charge and protonation state, or avoiding transformations that inappropriately break or form covalent bonds. (UNVERIFIED) has validated ligand selection criteria A relation that links an enhanced sampling process to the geometric or collective variable along which the system is steered or biased, for example in steered molecular dynamics or umbrella sampling. Examples are the distance between two groups of atoms, an angle, a dihedral angle, or a more complex function of atomic coordinates. has reaction coordinate A relation that links a molecular dynamics process to one of the discrete stages that make it up. The stages belong to the run rather than to the protocol, because one protocol may be followed by many runs. (UNVERIFIED) has simulation stage A relation that links a molecular dynamics process to a schedule defining how one of its parameters varies over simulated time or integration steps. A parameter that changes during a run, such as a temperature being raised or a restraint being released, is not described by a single value. has parameter schedule A relation that links a simulation stage to a continuity check assessing its temporal join with an adjacent stage. A long simulation is usually run as a chain of shorter stages, and a gap or an overlap between 2 of them makes the combined trajectory wrong. The check is what establishes that the stages really do join end to end. (UNVERIFIED) has continuity check A relation that links a simulation stage to an error encountered while loading or parsing one of its data files. Recording the error keeps a failed or partial read visible instead of silently producing incomplete metadata. A record with an error attached should not be treated as equivalent to one without. (UNVERIFIED) has parsing error A relation that links a molecular simulation system to a high-level classification label describing its composition. It lets systems be grouped and searched by what they contain rather than by their file names. The labels are coarse on purpose, so that a whole collection can be sorted without reading each entry. (UNVERIFIED) has system classification A relation that links a data format to the specific convention or schema variant that a file follows. One named format often covers several incompatible layouts, so naming the format alone does not say how to read a file. The convention is what a reader has to match. (UNVERIFIED) has format convention A relation that links a molecular dynamics process to the atom selection mask defining which atoms are subject to restraints. The mask is what turns "restrain the protein" into a definite set of atoms. A mask that selects the wrong atoms produces a simulation that runs normally and means something else. (UNVERIFIED) has restraint mask A relation that links a molecular dynamics process to a simulation parameter that governs how it is carried out, such as the integration timestep, the non-bonded cutoff, or an output write interval. These settings are part of what a result means, not incidental detail. A simulation cannot be repeated or judged without them. (UNVERIFIED) has simulation parameter A relation that links a molecular dynamics process to the simulation protocol it was carried out according to. The protocol is the reusable plan and the process is one execution of it, so a single protocol may be followed by many runs. (UNVERIFIED) follows protocol A relation that links a molecular simulation system to the water model used to represent its solvent, such as TIP3P or SPC/E. The choice fixes the geometry and the charges of every water molecule, and it affects properties that depend on the solvent, including diffusion and the strength of hydrogen bonds. (UNVERIFIED) has water model A relation that links a molecular simulation system to the chemical state chosen for the ends of its polymer chains, whether each end was closed off with a capping group or left uncapped and carrying a charge. The choice changes the total charge of the system, and it matters most for short peptides, where the two ends are a large part of the whole molecule. (UNVERIFIED) has capping group has termini state has specified input see is_input_of example_of_usage The inverse property of is specified input of 8/17/09: specified inputs of one process are not necessarily specified inputs of a larger process that it is part of. This is in contrast to how 'has participant' works. PERSON: Alan Ruttenberg PERSON: Bjoern Peters PERSON: Larry Hunter PERSON: Melanie Coutot has specified input is specified input of has specified output The inverse property of is specified output of PERSON: Alan Ruttenberg PERSON: Bjoern Peters PERSON: Larry Hunter PERSON: Melanie Courtot has specified output is specified output of A relation between a completely executed planned process and a continuant participating in that process. The presence of the continuant at the end of the process is explicitly specified in the objective specification which the process realizes the concretization of. Alan Ruttenberg PERSON:Bjoern Peters is specified output of a relation between a continuant and a process, in which the continuant is somehow involved in the process participates in A relationship between a generically dependent continuant and a specifically dependent continuant or process, in which the generically dependent continuant depends on some independent continuant or process in virtue of the fact that the specifically dependent continuant or process also depends on that same independent continuant. A generically dependent continuant may be concretized as multiple specifically dependent continuants or processes. is concretized as a relation between an independent continuant (the bearer) and a quality, in which the quality specifically depends on the bearer for its existence A bearer can have many qualities, and its qualities can exist for different periods of time, but none of its qualities can exist when the bearer does not exist. has quality p has input c iff: p is a process, c is a material entity, c is a participant in p, c is present at the start of p, and the state of c is modified during p. http://purl.obolibrary.org/obo/RO_0002233 has input p has output c iff c is a participant in p, c is present at the end of p, and c is not present in the same state at the beginning of p. has output has measurement value A relation that groups properties quantifying the chemical composition of a molecular system, such as the total number of atoms, residues and molecules of each type. Those counts give the stoichiometry and the overall size of the system being simulated. (UNVERIFIED) has composition count A relation that gives the total number of atoms in a molecular system, counting every molecule present including solvent and ions. It is the single figure that most determines what a simulation will cost, because the work of each step grows with it. A solvated protein system is usually tens of thousands of atoms, most of them water. (UNVERIFIED) NATOM The NATOM pointer in an AMBER prmtop topology file. number of total atoms NATOM A relation that gives the number of different atom types the force field uses for a system, rather than the number of atoms. An atom type is a category such as an aromatic carbon or a carbonyl oxygen, and every atom of that type shares the same interaction parameters. The figure reflects how chemically varied the system is. (UNVERIFIED) NTYPES The NTYPES pointer in an AMBER prmtop topology file. number of total distinct atoms NTYPES A relation that gives the total number of residues in a molecular system, a residue being 1 building block of a polymer, such as an amino acid, or a small whole molecule, such as water or an ion. In a system with explicit solvent most residues are usually water molecules, so the figure grows with the amount of solvent. (UNVERIFIED) NRES The NRES pointer in an AMBER prmtop topology file. number of total residues NRES A relation that gives the number of the last residue treated as part of the solute, which fixes where the receptor ends. Everything numbered beyond it is ligand, solvent or ions. The boundary matters because analyses of binding must know which atoms belong to the receptor. (UNVERIFIED) IPTRES The IPTRES pointer in an AMBER prmtop topology file. number of receptor residues IPTRES A relation that gives the number of ligand molecules in a molecular system. A ligand here is a small molecule being studied for how it binds to a larger one. No field in a topology file records this, so it has to be stated by whoever describes the system. (UNVERIFIED) number of ligands A relation that gives the number of ligand molecules in a system that carry a net electrical charge. It matters because charged ligands change how the system must be neutralised and how its electrostatics behave. No field in a topology file records this, so it has to be stated by whoever describes the system. (UNVERIFIED) number of charged ligands A relation that gives the number of distinct shapes of a ligand being considered in a study. Docking and conformational analysis generate many shapes of the same molecule, and how many were kept affects what the result means. No field in a topology file records this, so it has to be stated by whoever describes the system. (UNVERIFIED) number of ligand conformers A relation that gives the number of atoms in the biggest single residue of a system. A standard amino acid has fewer than 30, so a much larger figure marks something else, such as a cofactor, a ligand, a lipid or a modified residue. It is a quick way to notice that a system contains something other than a plain polymer. (UNVERIFIED) NMXRS The NMXRS pointer in an AMBER prmtop topology file. number of total atoms in the largest residue NMXRS A relation that gives the number of massless interaction sites in a molecular model, which carry charge or interaction parameters but sit at no nucleus. They let charge be placed where no atom is, such as on a lone pair of electrons or slightly off an oxygen. Water models with 4 sites and some halogen bonding parameters are the usual reason a system has any. (UNVERIFIED) NUMEXTRA The NUMEXTRA pointer in an AMBER prmtop topology file. number of extra points NUMEXTRA A relation that groups properties quantifying the covalent structure and connectivity of a molecular system, such as the number of bonds, angles and dihedral angles. Together they describe the topology, which is the graph of how the atoms are joined. (UNVERIFIED) has topological count A relation that gives the number of covalent bonds in a system in which at least 1 of the 2 atoms is a hydrogen. These bonds vibrate faster than anything else in the system, so they are usually held at a fixed length to permit a longer timestep. (UNVERIFIED) NBONH The NBONH pointer in an AMBER prmtop topology file. number of total bonds with hydrogens NBONH A relation that gives the number of covalent bonds in a system that join 2 heavy atoms, with no hydrogen involved. These bonds form the skeleton of each molecule and are not normally constrained, because they vibrate slowly enough for the timestep to follow them. (UNVERIFIED) MBONA The MBONA pointer in an AMBER prmtop topology file. number of total bonds without hydrogens MBONA A relation that gives the number of bond angles in a system in which at least one of the 3 atoms is a hydrogen. Angles involving hydrogen bend fastest, so some simulations constrain them as well as the bonds. Doing so permits a longer timestep still, at the cost of losing that motion. (UNVERIFIED) NTHETH The NTHETH pointer in an AMBER prmtop topology file. number of total angles with hydrogens NTHETH A relation that gives the number of bond angles in a system formed entirely by heavy atoms, that is atoms other than hydrogen. These angles set the shape of each molecule's skeleton and are almost never constrained. (UNVERIFIED) MTHETA The MTHETA pointer in an AMBER prmtop topology file. number of total angles without hydrogens MTHETA A relation that gives the number of dihedral angles in a system in which at least 1 of the 4 defining atoms is a hydrogen. A dihedral angle describes a twist defined by 4 atoms. (UNVERIFIED) NPHIH The NPHIH pointer in an AMBER prmtop topology file. number of total dihedral angles with hydrogens NPHIH A relation that gives the number of dihedral angles in a system defined entirely by heavy atoms, that is atoms other than hydrogen. They include the backbone dihedral angles of a protein, which largely control the shapes its chain can adopt. (UNVERIFIED) MPHIA The MPHIA pointer in an AMBER prmtop topology file. number of total dihedral angles without hydrogens MPHIA A relation that gives the number of atom pairs whose ordinary non-bonded interaction is deliberately left out of the calculation. Pairs joined through a bond, an angle or a dihedral are excluded because the bonded terms already describe how they interact, and in most force fields the pairs 3 bonds apart are then added back at reduced strength. The exclusion list is part of the topology rather than a setting. (UNVERIFIED) NNB The NNB pointer in an AMBER prmtop topology file. number of total excluded atom pairs NNB A relation that groups properties counting the degrees of freedom held fixed during a simulation, such as the bonds and angles made rigid by an algorithm like SHAKE. They record how far the natural dynamics of the system has been altered. (UNVERIFIED) has constraint count A relation that gives the number of covalent bonds whose length is held fixed by a constraint algorithm during a simulation. Each constrained bond removes 1 degree of freedom from the system, which is why the figure is needed to compute temperature correctly. (UNVERIFIED) prmtop NBONA - MBONA number of total bonds with constraints A relation that gives the number of bond angles held fixed by a constraint algorithm during a simulation. Constraining angles as well as bonds permits a longer timestep but removes the motion being constrained. (UNVERIFIED) prmtop NTHETA - MTHETA number of total angles with constraints A relation that gives the number of dihedral angles held fixed by a constraint algorithm during a simulation. Constraining a dihedral prevents a rotation entirely, so it is done rarely and deliberately. (UNVERIFIED) prmtop NPHIA - MPHIA number of total dihedrals with constraints A relation that groups properties counting the distinct force field parameters applied to a molecular system, such as the number of unique bond, angle and atom types. The counts indicate how complex the energy function used to model the system is. (UNVERIFIED) has force field parameter count A relation that gives the number of different bond parameter sets a system's topology draws on. Two bonds share a type when the force field gives them the same preferred length and stiffness. The figure reflects how chemically varied the bonding in the system is, not how many bonds there are. (UNVERIFIED) NUMBND The NUMBND pointer in an AMBER prmtop topology file. number of total unique bond types NUMBND A relation that gives the number of different angle parameter sets a system's topology draws on. Two angles share a type when the force field gives them the same preferred value and stiffness. The figure reflects the variety of local geometries the system contains. (UNVERIFIED) NUMANG The NUMANG pointer in an AMBER prmtop topology file. number of total unique angle types NUMANG A relation that gives the number of different dihedral parameter sets a system's topology draws on. Dihedral parameters control which rotations are preferred, so this figure reflects how varied the conformational behaviour of the system is. (UNVERIFIED) NPTRA The NPTRA pointer in an AMBER prmtop topology file. number of total unique dihedral types NPTRA A relation that gives the number of atom types defined in the force field parameter files being used, rather than the number appearing in a single system. It describes the coverage of the force field itself. (UNVERIFIED) NATYP The NATYP pointer in an AMBER prmtop topology file. number of total atom types in the parameter file NATYP A relation that gives the number of distinct pair types for the 10-12 hydrogen bond potential, an older treatment that gives hydrogen bonds an energy term of their own. Modern force fields describe hydrogen bonds through ordinary electrostatics and van der Waals terms instead, so this figure is 0 for most current systems. A non-zero value marks an older parameter set. (UNVERIFIED) NPHB The NPHB pointer in an AMBER prmtop topology file. number of total distinct hydrogen bond pair types NPHB A relation that groups properties counting what a chemical transformation in a free energy calculation covers, such as the atoms, bonds and angles being changed during the simulation. They set out the scope of the alchemical modification performed. (UNVERIFIED) has alchemical transformation count In an alchemical free energy calculation, this specifies the total count of bonds being chemically transformed. NBPER The NBPER pointer in an AMBER prmtop topology file. number of total perturbed bonds NBPER In an alchemical free energy calculation, this specifies the total count of angles being chemically transformed. NGPER The NGPER pointer in an AMBER prmtop topology file. number of total perturbed angles NGPER In an alchemical free energy calculation, this specifies the total count of dihedrals being chemically transformed. NDPER The NDPER pointer in an AMBER prmtop topology file. number of total perturbed dihedrals NDPER A relation that gives the number of bonds in an alchemical free energy calculation whose 2 atoms both lie in the group being transformed. An alchemical calculation changes 1 chemical group into another in small steps, to compute the free energy difference between them. (UNVERIFIED) MBPER The MBPER pointer in an AMBER prmtop topology file. number of total bonds with all atoms in the perturbed group MBPER A relation that gives the number of bond angles in an alchemical free energy calculation whose 3 atoms all lie in the group being transformed. An alchemical calculation changes 1 chemical group into another in small steps, to compute the free energy difference between them. (UNVERIFIED) MGPER The MGPER pointer in an AMBER prmtop topology file. number of total angles with all atoms in perturbed groups MGPER A relation that gives the number of dihedral angles in an alchemical free energy calculation whose 4 atoms all lie in the group being transformed. An alchemical calculation changes 1 chemical group into another in small steps, to compute the free energy difference between them. (UNVERIFIED) MDPER The MDPER pointer in an AMBER prmtop topology file. number of total dihedrals with all atoms in perturbed groups MDPER A relation that groups properties describing the overall state or the main qualitative features of a molecular system, such as yes-or-no flags and short textual descriptions of its composition and setup. They give high-level context about what kind of simulation it is. (UNVERIFIED) has state descriptor A relation that states whether a molecular system contains at least 1 ligand molecule. It marks at a glance which simulations concern binding rather than a molecule on its own. No field in a topology file records this, so it has to be stated by whoever describes the system. (UNVERIFIED) has ligand A relation that states whether salt ions were added to a system beyond the counterions needed to make it neutral. Added salt sets the ionic strength, which changes how strongly charges are screened and how molecules interact. No field in a topology file records this, so it has to be stated by whoever describes the system. (UNVERIFIED) has added salt A boolean descriptor indicating whether the molecular system houses an explicit lipid bilayer, rather than just an aqueous or implicit solvent environment. This is a higher-level descriptor not directly present in the prmtop pointers. is transmembrane system A relation that states whether the starting structure contained more than one arrangement for some part of the molecule. Crystal structures often record 2 positions for a flexible side chain, and which was used changes what the simulation began from. No field in a topology file records this, so it has to be stated by whoever describes the system. (UNVERIFIED) has multiple starting conformations A textual description of the types, proportions and spatial arrangements of different lipid molecules used to construct a biological membrane. This is a higher-level descriptor not directly present in the prmtop pointers. lipid composition A numerical or string representation of the relative amounts of different lipid types in a membrane. For example, a value of "3:1" could describe a mixture of two different lipids. This provides a semiquantitative description of the membrane composition. lipid ratio A relation that associates a molecular system with the index of the first solvent molecule in its ordered list of molecules. Molecules before it make up the solute, and molecules from it onwards make up the solvent. (UNVERIFIED) first solvent molecule index A relation that associates a molecular system with how many solvent molecules it contains. It quantifies the amount of solvent used to solvate the solute. (UNVERIFIED) number of solvent molecules A relation that associates a trajectory with the atom count recorded in each of its frames. It reflects the size of the coordinate set stored at every saved time point. (UNVERIFIED) number of atoms per frame A relation that associates a molecular system with a mapping from ion species to the count of ions of each species present. It records the ionic composition used, for example, to neutralize charge or set ionic strength. This property stores its value as text, not as a number. The text lists each kind of ion and how many of them there are, for example {"Na+": 12, "Cl-": 10}. That is how the simulation files themselves write it. The other count properties next to this one use plain numbers, because each of them counts a single thing. This one has to count several kinds of ion at the same time, and the ontology has no way yet to say which count belongs to which ion. We looked at two other ways of doing it: making a separate property for every kind of ion, or making a small object that holds one ion and its number. Both would add a lot of work for little gain right now, so we kept the text form. The text itself can be checked later with SHACL, a rule language for checking data files. ion counts dict A relation that associates a molecular system with how many bond-angle terms are defined in its force-field topology. It helps describe the size and complexity of the bonded interaction set. (UNVERIFIED) number of total angles A relation that associates a molecular system with how many dihedral, or torsion, terms are defined in its force-field topology. It helps describe the size and complexity of the bonded interaction set. (UNVERIFIED) number of total dihedrals A relation that associates a simulation with how many times a geometric constraint algorithm, such as SHAKE, failed to converge during the run. A nonzero value can signal instability in the integration. (UNVERIFIED) number of constraint failures A relation that associates a simulation run with an indication of whether it finished normally or terminated abnormally. It supports quick screening of runs for successful completion. (UNVERIFIED) has completion status A relation that indicates whether a run was started as a continuation of a previous run rather than from initial conditions. It distinguishes restarted runs from fresh ones. (UNVERIFIED) continuation flag A relation that indicates whether the atomic coordinates in a trajectory have been translated back into the primary periodic unit cell. It records whether periodic-image wrapping was applied to the stored coordinates. (UNVERIFIED) coordinate wrapping flag A relation that indicates whether a simulation uses a combined quantum-mechanics and molecular-mechanics treatment for part of the system. It records whether a region is described at the quantum level while the remainder uses a classical force field. (UNVERIFIED) QM/MM activation flag A relation that allows the software to apply a time-averaged or spatial-averaged guiding force derived from the system's own trajectory to accelerate low-frequency, concerted motions and enhance conformational sampling. self-guided Langevin dynamics flag A relation that controls whether a spherical portion of solvent (water) is applied around a solute, restrained by a soft half-harmonic potential. spherical cap flag A relation that indicates whether periodic box dimension information is present for a molecular system or trajectory. It records whether unit-cell size and shape data are available. (UNVERIFIED) has box dimensions flag A relation that indicates whether atomic velocity data are stored alongside coordinates in a file or trajectory. Velocities are needed to continue a run exactly from where it stopped. (UNVERIFIED) has velocities flag OBSOLETE. A relation that records the kind of periodic box used for a molecular system, such as rectangular or truncated octahedral. It identifies the geometry of the periodic unit cell. Retired 2026-08-13. It duplicated MOLSIM:000297 has periodic box type, which links a molecular simulation system to the class MOLSIM:000032 box type and its five children. This term held the same information as a plain string, so no value could be checked, while the object property points at a real hierarchy that a reasoner can use. Its label also broke the house convention: ten of the eleven data properties ending in flag are yes-or-no values, and this one recorded which box geometry was used. obsolete box type flag true A relation that groups properties describing a molecular dynamics trajectory as a whole, such as how many frames it holds and the format its coordinates are stored in. (UNVERIFIED) Domain chosen 2026-08-13. This domain asserts something the other grouping properties here do not: anything carrying a child of this property, such as MOLSIM:002138 number of frames, is inferred to be a MOLSIM:001081 simulation trajectory. That is correct if a frame count is only ever recorded about a trajectory. If the source metadata ever records a frame count against a whole run rather than against its trajectory file, the domain should be MOLSIM:000000 molecular dynamics process instead. Confirm against the MDDB records before relying on it. has trajectory metadata A relation that associates a trajectory with how many saved frames it contains. It reflects how many time points were recorded over the course of the run. (UNVERIFIED) number of frames A relation that associates a trajectory or coordinate file with an identifier of the file format used to store its coordinates. It records how the coordinate data are encoded, for example as NetCDF or formatted text. (UNVERIFIED) coordinate format specifier A relation that groups properties whose values are ordered lists of numbers describing the size and shape of the periodic simulation box of a molecular system, such as its edge lengths and angles. Each list is stored as text. (UNVERIFIED) has box geometry array A relation that associates a molecular system with an array giving the lengths and angles that define its periodic simulation box. It captures the full geometry of the periodic unit cell. (UNVERIFIED) box dimensions array A relation that associates a molecular system with an array giving the periodic box lengths and angles at the start of a run. It records the initial unit-cell geometry before any volume changes. (UNVERIFIED) initial box dimensions array A relation that groups properties recording the names and labels assigned to components of a molecular system, such as residue and atom-type names. (UNVERIFIED) has system nomenclature A relation that associates a molecular system with the residue name used to identify its solvent in the topology, such as WAT for water. It records the label under which solvent is stored. (UNVERIFIED) solvent residue name A relation that associates a molecular system with the ordered list of residue names for its constituent residues. It records the per-residue naming as given in the system's topology. (UNVERIFIED) residue labels A relation that associates a molecular system with the set of distinct residue names it contains. It summarizes the kinds of residues present without regard to how many times each occurs. (UNVERIFIED) unique residue names A relation that associates a molecular system with the set of distinct force-field atom types it uses. It summarizes the range of parameterized atom types present in the topology. (UNVERIFIED) unique atom types A relation that associates a molecular system with the identifier of the intrinsic radii set used in a generalized Born implicit-solvent calculation. It records which parameterization defines the atomic solvation radii. (UNVERIFIED) born radii set name A relation that groups properties recording what a simulation or a checking step reported about itself, such as an error message, a warning, or the outcome of a validation check. It gathers the output a run produced about its own progress, as distinct from the settings it was given. (UNVERIFIED) has diagnostic record A relation that associates a simulation or processing step with textual diagnostic output emitted when a fault occurred. It captures the message describing why a run failed. (UNVERIFIED) has error message A relation that associates a simulation or processing step with a textual warning emitted during execution. Such warnings flag potential problems that did not necessarily stop the run. (UNVERIFIED) has warning message A relation that associates a dataset or run with a non-fatal issue raised during a validation check. It flags a concern found while verifying correctness or consistency. (UNVERIFIED) has validation warning A relation that associates a dataset or run with a fault raised during a validation check that caused the check to fail. It records a validation failure. (UNVERIFIED) has validation error A relation that associates a dataset or run with the outcome of a check verifying that its data are mutually consistent. It records whether internal consistency tests passed or failed. (UNVERIFIED) consistency check result A relation that indicates whether a trajectory or restart file contains atomic coordinate data. Almost every trajectory does, but a file holding only velocities or only forces does not, and reading one expecting coordinates fails. The flag lets a tool decide what a file is good for before opening it. (UNVERIFIED) has coordinates array A relation that indicates whether a trajectory or restart file contains per-atom force data. Forces are written less often than coordinates, because they add to the size of a trajectory and are needed only for some uses, such as force analysis or training machine-learning models. (UNVERIFIED) has forces array A relation that indicates whether a trajectory file contains per-frame time-stamp data. Without it the frames have an order but no times, so any quantity involving a rate has to assume the interval between them. Stored times also reveal a gap where frames were lost. (UNVERIFIED) has time array A relation that indicates whether a simulation applies NMR-derived restraints or dynamically varying conditions. Such restraints pull the system towards experimental measurements, so the resulting structures reflect the data as well as the force field. A trajectory produced this way must not be treated as a free simulation. (UNVERIFIED) has NMR restraints active A relation that indicates whether metadata extraction ran in a degraded mode because one or more input files were missing or unreadable. A record produced this way is incomplete rather than wrong, and the missing fields are absent rather than zero. Marking it prevents a partial record being read as a complete one. (UNVERIFIED) is degraded extraction mode A relation that indicates whether a run applies a time-varying target-temperature schedule. A run that heats or cools during its course is not described by a single temperature, so any average taken across it would be meaningless. The flag warns that the run is not at a steady state. has temperature schedule A relation that indicates whether a run applies a time-varying positional-restraint weight schedule. Restraints are commonly released gradually while a system settles, so the strength recorded at the start does not describe the whole run. The flag marks a run in which the restraints changed. (UNVERIFIED) has restraint schedule A relation that indicates whether a run applies a time-varying non-bonded cutoff schedule. Changing the cutoff during a run changes how much of each interaction is included, so energies from different parts of the trajectory are not comparable. It is unusual, which is why it is worth recording when it happens. (UNVERIFIED) has cutoff schedule A relation that associates a molecular system with a descriptive feature or heuristic modification of its force field, such as CMAP terms, which correct backbone dihedral energies, or conversion from CHARMM to AMBER format by the CHAMBER tool. These details are not conveyed by the name of the force field, so recording them keeps the force field identifiable from the metadata alone. (UNVERIFIED) force field feature A relation that associates a molecular system with a serialized mapping from each residue type to the number of residues of that type present. It says what the system is made of in one value, without a separate term for every residue type. It is stored as text because the set of residue types is open-ended. (UNVERIFIED) residue composition map A relation that associates a simulation with the complete, serialized set of key-value control parameters parsed from its input file. It keeps every setting as written, so nothing in the original input is lost. It is a record for a person to read rather than a structure for a machine to reason over. (UNVERIFIED) raw control parameters A relation that determines specialized algorithmic modifications designed to bypass the steep computational cost of standard REMD by applying enhanced sampling to only a specific subsystem or localized region. REMD variant type A relation that associates a simulation stage with the file path or identifier of the restart file used to initiate it. It is what connects one stage to the one before it, so a chain of stages can be reconstructed after the fact. Without it the stages of a long run are a set of files in no particular order. (UNVERIFIED) restart file path A relation that indicates whether the total electrostatic charge of a molecular system is effectively zero, that is, whether the system is electrically neutral. A charged periodic system interacts with its own images in a way that distorts energies, so counter-ions are normally added to neutralise it. A system reported as not neutral needs that explained. (UNVERIFIED) is charge neutral A relation that associates a molecular system with a short human-readable summary of the hydrogen masses observed, such as the mass range and the count of hydrogen atoms, used to describe hydrogen mass repartitioning. Repartitioning moves mass from heavy atoms onto the hydrogens bonded to them, which slows the fastest vibrations and permits a longer timestep. The summary shows at a glance whether it was applied and how far. (UNVERIFIED) hydrogen mass summary A relation that associates a graphics processing unit with the model name of the specific device used to run a simulation, such as the name reported by the CUDA runtime. Recording the device helps explain the speed of a run and makes it possible to repeat the run on the same hardware. (UNVERIFIED) gpu model name A relation that indicates whether hydrogen mass repartitioning was inferred from the integration timestep, specifically a timestep of 3 femtoseconds or larger, rather than detected directly from the hydrogen masses in the topology. The inference is a reasonable guess and not a reading, because such a timestep is otherwise unstable. HMR inferred from timestep flag A relation that associates a simulation protocol with the total number of integration steps performed across all of its stages. When every stage uses the same timestep, multiplying the 2 gives the total simulated time. (UNVERIFIED) total simulation steps A relation that associates a molecular system with the number of atoms belonging to its proteins. Proteins are the large chain-like molecules built from amino acids; this reports how many atoms make up all of them together. (UNVERIFIED) protein atom count A relation that associates a molecular system with the number of atoms belonging to its nucleic acids, that is DNA and RNA taken together. It separates the nucleic acid part of a system from protein, solvent and ions in a single figure. A value of 0 marks a system with no nucleic acid at all. (UNVERIFIED) nucleic acid atom count A relation that associates a molecular system with the number of atoms belonging to its lipids. Lipids are the fatty molecules that form cell membranes. (UNVERIFIED) lipid atom count A relation that associates a molecular system with the number of atoms belonging to ions. In most simulations these ions are single charged atoms, such as sodium or chloride, added to balance the charge or set the salt concentration. (UNVERIFIED) ion atom count A relation that associates a molecular system with the number of atoms belonging to its solvent, which is usually water. It counts atoms, not whole solvent molecules. (UNVERIFIED) solvent atom count A relation that associates a molecular system with the number of atoms that fall into none of the named categories (protein, nucleic acid, lipid, ion, solvent). It captures leftovers such as small-molecule ligands or cofactors. (UNVERIFIED) other atom count A relation that associates a molecular system with the number of residues in its proteins. A residue is one amino-acid building block of a protein chain. (UNVERIFIED) protein residue count A relation that associates a molecular system with the number of residues in its nucleic acids. Here a residue is one nucleotide building block of a DNA or RNA chain. (UNVERIFIED) nucleic acid residue count A relation that associates a molecular system with the number of lipid residues it contains. Some force fields write each lipid molecule as 1 residue, while others, such as the AMBER LIPID21 force field, split it into a head group residue and separate tail residues, so this figure need not equal the number of lipid molecules. (UNVERIFIED) lipid residue count A relation that associates a molecular system with how many nucleosomes it contains. A nucleosome is the basic repeating unit of chromatin, a stretch of DNA wound around a core of histone proteins. (UNVERIFIED) number of nucleosomes A relation that associates the system with the length of a DNA strand, measured as the number of nucleotide bases it contains. Because a system can contain several strands, this may be recorded once per strand. (UNVERIFIED) number of nucleotides per DNA strand A relation that associates a molecular system with an ordered list of the 3 angles between the edges of its periodic simulation box, usually written α, β and γ. The list is stored as text, as read from or written to a simulation file. (UNVERIFIED) box angles array A relation that states whether a saved trajectory includes the atomic positions (the 3-D coordinates of every atom over time). Positions are the most basic trajectory data and are almost always present. (UNVERIFIED) has positions flag A relation that states whether a saved trajectory includes the forces acting on each atom over time. Forces are optional extra data, useful for example when training machine-learning models. (UNVERIFIED) has forces flag A relation that records the version number of a piece of software, such as an MD engine, the program that runs the simulation (for example, version 2024.1). Recording the version helps make a simulation reproducible, because software behaviour can change between versions. (UNVERIFIED) software version A relation that records whether a simulated system includes an antibody, a Y-shaped immune protein that recognises and sticks to a specific target such as a virus. (UNVERIFIED) contains antibody A relation that records whether a simulated system includes a nanobody, a very small, single-piece version of an antibody, originally from animals such as llamas, used as a compact binding tool. A nanobody is a single antibody domain, roughly a tenth the size of a full antibody. (UNVERIFIED) contains nanobody A relation that records whether the system includes a linker histone, such as histone H1, an extra histone that binds the DNA where it enters and leaves the nucleosome core. It is not part of the core of 8 histone proteins and is missing from many experimental structures, so its presence changes what a simulation represents. (UNVERIFIED) contains linker histone A relation that records whether the histone proteins in a system still have their tails, the flexible ends that extend out from the nucleosome core and are often chemically modified to control gene activity. The tails have no fixed shape, so they are often removed from a model to make it easier to simulate, which also removes the parts that carry those chemical marks. (UNVERIFIED) histone tails present A relation that groups properties recording what a simulation run was given to start from, such as the settings parsed from its input file or the restart file it continued from. It covers the input side of a run, as distinct from what the run reported about itself. (UNVERIFIED) has run input record A relation that associates a molecular dynamics process with the file path or identifier of the topology file defining the simulated system, such as an AMBER prmtop file or a GROMACS top file. The topology file states which atoms are present, how they are bonded, which parameters they were assigned, and which chemical states were assumed when the system was built. (UNVERIFIED) MOLSIM records the composition of a simulated system at the system level and points at the topology and coordinate files for the per-residue detail, rather than restating that detail term by term. This is why there is no vocabulary for individual histidine or cysteine states: those are decided per residue and per atom, and the topology file states them exactly. Paired with MOLSIM:002257 has coordinate file path, which records the starting positions. has topology file path A relation that associates a molecular dynamics process with the file path or identifier of the coordinate file holding the starting position of every atom, such as an AMBER inpcrd or rst7 file or a GROMACS gro file. Together with the topology file it fixes the exact system state a run began from. (UNVERIFIED) has coordinate file path A relation that associates a molecular simulation system with any solvent added alongside water, such as urea, glycerol or an alcohol. A cosolvent changes how the solute behaves, and a record that omits it describes a different system from the one that was run. (UNVERIFIED) Holds text rather than pointing at a named class for each substance, and that is a deliberate choice rather than a shortcut. Pointing at named classes would need MOLSIM classes for urea, glycerol, DMSO and every other cosolvent, which is a chemistry vocabulary that ChEBI already provides with roughly 190,000 terms. The ChEBI import was removed on 2026-08-24 because all 189 classes it delivered were unreferenced, and the agreed approach records what a system is made of at the system level while pointing at the topology file for the per-residue chemical detail. The cost of the text form is real: urea, Urea, URE and 8M urea are the same substance and different strings, so a search by substance is unreliable. Revisit only when a specific record must be queried by substance rather than read by a person. The fix then is to import the few named ChEBI terms actually needed, bounded so that no ancestors arrive with them, and change this property to point at those. The same reasoning applies to the data properties MOLSIM:001161 lipid composition and MOLSIM:001162 lipid ratio, which also hold text. This property stores its value as text, because a system may contain more than 1 cosolvent and each may carry a concentration. Values look like "urea 8 M" or "glycerol 30 percent, DMSO 5 percent". Checking the text follows a fixed shape is a job for SHACL rather than OWL. has cosolvent A relation that records whether the 2 layers of a simulated membrane hold the same mixture of lipids as each other. Real cell membranes usually differ between their 2 sides, so whether a simulation copied that difference or ignored it changes what the results can be said to show. (UNVERIFIED) has symmetric leaflet composition A relation that associates a molecular simulation system with the size of its membrane patch in the plane of the membrane. The patch must be wide enough that a protein does not interact with its own periodic copy across the boundary, so the size limits which conclusions the simulation can support. (UNVERIFIED) This property stores its value as text, because the size has 2 numbers and a unit. Values look like "9.5 x 9.5 nm" or "120 x 95 angstrom". It carries no unit restriction for the same reason MOLSIM:001213 periodic box dimensions carries none: the unit and value pattern describes a single number, and a patch size is two. A single length in this domain does take that pattern, as MOLSIM:001875 membrane width does. Checking the text follows a fixed shape is a job for SHACL rather than OWL. has membrane dimensions A relation that associates a molecular simulation system with the rule used to delete lipids that overlapped the protein when the membrane was built. Placing a protein into a lipid patch always leaves some lipids sitting where the protein is, and the rule chosen for removing them decides how many lipids the finished system holds. (UNVERIFIED) This property stores its value as text, because the rule is usually a short description rather than a single number. Values look like "lipids with any atom within 0.6 nm of the protein were deleted" or "removed by the InflateGRO procedure". Checking the text follows a fixed shape is a job for SHACL rather than OWL. has lipid clash removal criterion A complex comprised of DNA wound around a multisubunit core and associated proteins, which forms the primary packing unit of DNA into higher order structures. GO:0000787 GO:0000788 GO:0005718 Wikipedia:Nucleosome cytoplasmic nucleosome nuclear nucleosome cellular_component GO:0000786 nucleosome A complex comprised of DNA wound around a multisubunit core and associated proteins, which forms the primary packing unit of DNA into higher order structures. GOC:elh journal article Examples are articles published in the journals, Nature and Science. The content can often be cited by reference to a paper based encoding, e.g. Authors, Title of article, Journal name, date or year of publication, volume and page number. A report that is published in a journal. PERSON: Alan Ruttenberg PERSON: Chris Stoeckert OBI_0000159 group:OBI journal article data entity information content entity Examples of information content entites include journal articles, data, graphical layouts, and graphs. A generically dependent continuant that is about some thing. 2014-03-10: The use of "thing" is intended to be general enough to include universals and configurations (see https://groups.google.com/d/msg/information-ontology/GBxvYZCk1oc/-L6B5fSBBTQJ). information_content_entity 'is_encoded_in' some digital_entity in obi before split (040907). information_content_entity 'is_encoded_in' some physical_document in obi before split (040907). Previous. An information content entity is a non-realizable information entity that 'is encoded in' some digital or physical entity. PERSON: Chris Stoeckert OBI_0000142 information content entity The curation status of the term. The allowed values come from an enumerated list of predefined terms. See the specification of these instances for more detailed definitions of each enumerated value. GROUP:OBI:<http://purl.obolibrary.org/obo/obi> OBI_0000266 curation status specification report Examples of reports are gene lists and investigation reports. These are not published (journal) articles but may be included in a journal article. A document assembled by an author for the purpose of providing information for the audience. A report is the output of a documenting process and has the objective to be consumed by a specific audience. Topic of the report is on something that has completed. A report is not a single figure. Examples of reports are journal article, patent application, grant progress report, case report (not patient record). 2009-03-16: comment from Darren Natale: I am slightly uneasy with the sentence "Topic of the report is on something that has completed." Should it be restricted to those things that are completed? For example, a progress report is (usually) about something that definitely has *not* been completed, or may include (only) projections. I think the definition would not suffer if the whole sentence is deleted. 2009-03-16: this was report of results with definition: A report is a narrative object that is a formal statement of the results of an investigation, or of any matter on which definite information is required, made by some person or body instructed or required to do so. 2009-03-16: work has been done on this term during during the OBI workshop winter 2009 and the current definition was considered acceptable for use in OBI. If there is a need to modify this definition please notify OBI. 2009-08-10 Alan Ruttenberg: Larry Hunter suggests that this be obsoleted and replaced by 'document'. Alan restored as there are OBI dependencies and this merits further discussion disagreement about where reports go. alan: only some gene lists are reports. Is a report all the content of some document? The example of usage suggests that a report may be part of some article. Term needs clarification PERSON: Alan Ruttenberg PERSON: Chris Stoeckert PERSON: Melanie Courtot GROUP: OBI OBI_0000099 report Data about an ontology part is a data item about a part of an ontology, for example a term data about an ontology part plan specification PMID: 18323827.Nat Med. 2008 Mar;14(3):226.New plan proposed to help resolve conflicting medical advice. A directive information entity with action specifications and objective specifications as parts, and that may be concretized as a realizable entity that, if realized, is realized in a process in which the bearer tries to achieve the objectives by taking the actions specified. 2009-03-16: provenance: a term a plan was proposed for OBI (OBI_0000344) , edited by the PlanAndPlannedProcess branch. Original definition was " a plan is a specification of a process that is realized by an actor to achieve the objective specified as part of the plan". It has been subsequently moved to IAO where the objective for which the original term was defined was satisfied with the definitionof this, different, term. 2014-03-31: A plan specification can have other parts, such as conditional specifications. 2022-01-16 Updated definition to that proposed by Clint Dowloand, IAO Issue 231. Alternative previous definition: a plan is a set of instructions that specify how an objective should be achieved Alan Ruttenberg Clint Dowland OBI Plan and Planned Process branch OBI_0000344 2/3/2009 Comment from OBI review. Action specification not well enough specified. Conditional specification not well enough specified. Question whether all plan specifications have objective specifications. Request that IAO either clarify these or change definitions not to use them plan specification https://github.com/information-artifact-ontology/IAO/issues/231#issuecomment-1010455131 The reason for which a term has been deprecated. The allowed values come from an enumerated list of predefined terms. See the specification of these instances for more detailed definitions of each enumerated value. obsolescence reason specification document A journal article, patent application, laboratory notebook, or a book A collection of information content entities intended to be understood together as a whole PERSON: Lawrence Hunter document publication journal article, newspaper story, book, etc. A document that is the output of a publishing process. PERSON: Chris Stoeckert PERSON: Jie Zheng PERSON: Lawrence Hunter published document Revisit the term in Octorber 2020. Improve the defintion. publication A denotator type indicates how a term should be interpreted from an ontological perspective. Barry Smith, Werner Ceusters denotator type A sequence of related actions or operations performed to achieve a specific outcome in a biomolecular simulation study. This encompasses all stages from initial system setup and simulation execution to final data analysis and interpretation. molecular dynamics process A set of instructions and data that a computer follows to carry out a task. In molecular simulation it covers everything from a program that runs a simulation to a small script that converts one file format into another. (UNVERIFIED) software A stage of a simulation study in which the molecular system and its environment are assembled and made ready to run. It covers obtaining a structure, adding the atoms and molecules the experimental structure left out, choosing a force field, adding solvent and ions, and bringing the system to equilibrium before data is collected. (UNVERIFIED) preparation simulation preparation The stage of a biomolecular simulation process where the main simulation runs are performed according to the prepared system and defined parameters to generate trajectory data. This phase involves propagating the system forward in time using algorithms like molecular dynamics or Monte Carlo, under specific ensemble conditions. The goal is typically to sample conformational space, observe dynamic events, or calculate specific properties like free energies. execution simulation execution A stage of a simulation study that follows the run, in which the trajectory and energy output are processed into results. It covers everything from measuring a distance to fitting a kinetic model. It turns the raw output of a simulation into scientific findings. (UNVERIFIED) analysis simulation analysis A simulation preparation stage in which the ionization states of ionizable groups in residues and ligands, that is which protons each carries, are assigned from their estimated pKa values at the target pH of the simulation. Correct assignment matters because it sets the charges, and so the electrostatic interactions, of the model, and any pH-dependent behavior. Specialized tools are often used for this task. (UNVERIFIED) protonation A simplified representation used to study a real-world system, such as a molecule or collection of molecules. In biomolecular simulation, it defines the components (atoms, residues) and their properties needed for computational analysis. This framework allows scientists to simulate and predict the system's behavior. (UNVERIFIED) Distinct from 'MoDEL' (MOLSIM:001001), a named simulation data repository whose name differs from this term only by capitalisation. This term is the general class of representations used in simulation. model A set of mathematical functions and fitted numbers used to compute the potential energy of a system of atoms from their positions alone. Instead of treating electrons explicitly, most such models use simple terms for bonds, angles, rotations about bonds and non-bonded contacts. This choice is a main factor in how accurate a classical simulation is. (UNVERIFIED) FF force field A force field specifically designed and parameterized for simulating lipid molecules and lipid bilayers, crucial components of cell membranes. It aims to accurately capture the unique properties of lipids, including their phase behavior and interactions with other molecules. Examples include the AMBER Lipid14, Lipid17 and Lipid21 parameter sets and the lipid parameters of CHARMM36. (UNVERIFIED) lipid force field A force field specifically designed and parameterized for simulating proteins and peptides, focusing on accurately representing amino acid properties and polypeptide chain behavior. These are crucial for studying protein folding, dynamics, and interactions. Many different protein force fields exist, often differing in their parameterization philosophy and target data. (UNVERIFIED) protein force field A force field specifically developed and parameterized for simulating Deoxyribonucleic Acid (DNA) molecules. It focuses on accurately representing the structure, dynamics, and interactions of DNA bases, sugars, and phosphate backbone, including features like sugar pucker and helical parameters. Accurate DNA force fields are essential for studying DNA replication, transcription, and drug binding. DNA force field A force field specifically developed and parameterized for simulating Ribonucleic Acid (RNA) molecules. It aims to accurately capture the complex conformational landscape and catalytic potential of RNA. Development often focuses on improving descriptions of sugar pucker, backbone torsions (especially the epsilon and zeta angles), and non-canonical base interactions. RNA force field A force field specifically designed and parameterized for simulating carbohydrate molecules (sugars or glycans). It addresses the unique challenges of modeling the flexibility of glycosidic linkages between sugar units and the numerous hydrogen bonding possibilities involving hydroxyl groups. These are important for studying glycoproteins, glycolipids, and polysaccharides like cellulose. carbohydrates force field A force field parameterised for small organic molecules, covering the wide range of chemistry found in drug-like compounds. It must handle groups no protein or nucleic acid force field was built for. It is used together with a biomolecular force field when a ligand is simulated in a protein. (UNVERIFIED) organic molecule force field A force field parameterised for water alone, usually representing a molecule by 3 to 5 interaction points with fixed geometry. Water is most of the atoms in a biomolecular simulation, so the choice affects cost as much as accuracy. No model reproduces every measured property of water, so the choice depends on which properties matter for the study. (UNVERIFIED) water model force field A set of parameters specifically developed to describe the interactions of ions (charged atoms or small molecules) within a simulation, typically in conjunction with a specific water model and other biomolecular force fields. These parameters define the ion's charge and non-bonded terms (usually Lennard-Jones parameters for size and interaction strength). Accurate ion parameters are crucial for modeling salt effects, ion channels, and metalloproteins, and often require careful tuning for compatibility with the chosen water model. (UNVERIFIED) ion force field A chemical entity-based force field designed to give parameters for almost any organic molecule, such as a drug or a ligand, rather than for 1 class of biomolecule. It achieves that breadth by assigning atom types from rules based on each atom's chemical environment, and by supplying rules to estimate parameters that are missing. The parameters it gives for a particular molecule are approximate, so they are often checked or refined. (UNVERIFIED) other organic molecules force field general purpose organic force field A basis set that uses a single function for each core and valence atomic orbital of an atom, with no extra functions. Because each atomic orbital is described by 1 function of fixed size, it cannot expand or contract when a bond forms, so bonding is described poorly. The STO-nG sets are common examples; they are cheap and used mainly for teaching and very rough calculations. (UNVERIFIED) minimal minimal basis set A lipid force field for the AMBER family that builds on Lipid14 and Lipid17 and extends the modular phospholipid residues to anionic head groups, polyunsaturated tails and sphingomyelin. Its bonded alkane parameters were refitted to quantum energies, and all head groups were given new partial charges. It is the lipid force field AMBER currently recommends. (UNVERIFIED) lipid21 A lipid force field for the AMBER family, released in 2014, that builds on the modular head group and tail scheme of Lipid11. It allowed lipid bilayers to be simulated at constant pressure without an artificial surface tension term, a major advance over earlier AMBER lipid parameters. It was later extended with cholesterol parameters and by Lipid17 and Lipid21. (UNVERIFIED) lipid14 A specific lipid force field parameter set developed for the AMBER simulation package, released around 2017. It was built upon the modularity of LIPID14 and provided an extension of modular phospholipid residues to include anionic head groups and polyunsaturated tails. lipid17 A specific protein force field parameter set from the AMBER family (ff), released around 2014, with optimizations primarily for side-chain (S) and backbone (B) torsion angles compared to earlier versions like ff99SB. It was widely adopted for its improved accuracy in representing protein secondary structure and dynamics. It is an all-atom force field. AMBER ff14SB A rigid water model with 3 interaction sites, 1 on each atom, using the measured geometry of an isolated water molecule. It is the water model the AMBER force fields were developed with, and the CHARMM force fields use a modified form with extra Lennard-Jones terms on hydrogen. It reproduces the density of water well, but water molecules diffuse about twice as fast as in real water, which matters for any rate taken from a simulation. (UNVERIFIED) TIP3P A force field for small organic molecules, built to be used alongside the AMBER protein and nucleic acid parameters. It defines atom types broadly enough to cover almost any organic structure, and supplies rules for filling gaps where no parameter was fitted. That breadth is why its parameters for any particular molecule should be checked rather than trusted. (UNVERIFIED) GAFF general AMBER force field A force field for small organic molecules that is the second version of the general AMBER force field, with parameters refitted against a larger set of quantum calculations and measurements than the first. It is the general force field AMBER recommends for ligands, and is meant to be combined with RESP or ABCG2 charges rather than with AM1-BCC charges. Its parameters differ from those of the first version, so the 2 are not interchangeable. (UNVERIFIED) GAFF2 general AMBER force field 2 A text molecular structure format that extends the Tripos mol2 format with extra records naming the atoms through which a residue links to its neighbours. LEaP in AmberTools can load it, so that a residue built from it can be joined into a larger molecule without the links being defined again by hand. (UNVERIFIED) mol3 format A method for processing the data a simulation produces and extracting something meaningful from it. It turns a trajectory, which is a list of coordinates at successive times, into a quantity that can be compared with an experiment or with another simulation. Almost nothing useful comes out of a simulation without one. (UNVERIFIED) analysis method A specific type of analysis method focused on understanding how chemical reactions occur within the simulated molecular system. It involves characterizing the pathways, transition states, and energy changes associated with the transformation of reactants into products. This helps elucidate reaction mechanisms and kinetics at the molecular level. reaction analysis A reaction analysis that uses the positions (coordinates) of atoms to describe the progress of a chemical reaction or conformational change. It defines one or more geometric variables, collectively known as reaction coordinates, to map the transition pathway between different molecular states. This analysis helps visualize and quantify the structural changes during a reaction. (UNVERIFIED) coordinate analysis A type of coordinate analysis where the reaction progress is described using simple, intuitive geometric parameters like bond distances, angles, or dihedrals directly involved in the transformation. This provides a straightforward way to monitor specific structural changes assumed to be representative of the reaction pathway. direct reaction coordinate analysis OBSOLETE. Refers to specific parameterizations or types of water models commonly used in biomolecular simulations, often used somewhat interchangeably with 'water model'. Each 'water type' (e.g., SPC, TIP3P, TIP4P-Ew) defines the geometry, number of interaction sites, and associated parameters (charges, Lennard-Jones) for representing water molecules. The choice significantly influences simulation outcomes, particularly solvation effects and dynamics. Retired 2026-08-24. Its own definition described it as 'often used somewhat interchangeably with water model', and that is what it was: a second home for water models alongside MOLSIM:000014 water model force field, which holds all 18. Its two children were re-parented rather than replaced, so there is no single successor. MOLSIM:000031 optimal point charges moved to MOLSIM:000014 beside OPC3, and MOLSIM:000476 TIP3P water model was retired as a duplicate of MOLSIM:000022 TIP3P. obsolete water type true A rigid four-site water model whose partial charges and geometry were optimised together to reproduce bulk properties of liquid water, including the density, the heat of vaporisation and the dielectric constant. It reproduces several of those properties more closely than the older three-site and four-site models at a similar computational cost. (UNVERIFIED) optimal point charges OPC Specifies the geometric shape of the simulation cell used under periodic boundary conditions (PBCs), which replicate the central simulation system in three dimensions to mimic a bulk environment and avoid surface effects. The choice of shape (e.g., cubic, truncated octahedron, rhombic dodecahedron) affects the computational efficiency (volume of solvent needed) and how well the unit cell packs space. It determines the distance and geometry for interactions between a particle and its periodic images. box type A specific type of simulation box shape corresponding to a dodecahedron with rhombus-shaped faces. It is one of the space-filling polyhedra that is closer to a sphere than a cube, thus often requiring less solvent volume (and fewer solvent molecules) to solvate a roughly spherical solute while maintaining a minimum distance to the periodic images. It can offer computational savings compared to a cubic box of similar minimum image distance. rhombic dodecahedron A specific type of simulation box shape created by cutting off the corners of an octahedron. Like the rhombic dodecahedron, it is a space-filling polyhedron that approximates a sphere more closely than a cube, reducing the required solvent volume for solvating non-cubic solutes under periodic boundary conditions. It offers similar computational efficiency benefits as the rhombic dodecahedron for many systems. truncated octahedron A simulation box shape consisting of a hexagonal base extruded along a perpendicular axis. This shape is sometimes used for systems with inherent one-dimensional periodicity or symmetry, such as fibers (like DNA) or layered systems, where periodic boundary conditions are naturally applied across the hexagonal faces and along the prism axis. Its use is less common than cubic or truncated octahedral boxes for general globular protein simulations. hexagonal prism The simplest simulation box shape, defined by three mutually perpendicular axes (orthogonal box). If all edge lengths are equal, it is cubic; otherwise, it is rectangular (also called orthorhombic or cuboid). While straightforward to implement and visualize, cubic boxes can be inefficient for solvating non-cubic solutes, often requiring significantly more solvent molecules than more spherically approximating shapes like the truncated octahedron to achieve the same minimum distance between periodic images. rectangular/cubic A well-defined sequence of steps or rules designed to perform a specific task or solve a particular problem. In computational science, algorithms provide the precise instructions a computer follows to reach a desired result, such as simulating molecular motion or analyzing data. (UNVERIFIED) algorithm An algorithm used in molecular dynamics simulations to control the temperature of the simulated system, ensuring it samples configurations corresponding to a specific target temperature (e.g., for the NVT or NPT ensembles). Thermostats work by adjusting particle velocities or adding/removing kinetic energy in various ways to mimic coupling with an external heat bath. Different thermostat algorithms vary in their theoretical basis, efficiency, and how accurately they reproduce correct statistical mechanics. (UNVERIFIED) thermostat algorithm An algorithm used in molecular dynamics simulations, typically alongside a thermostat, to control the pressure of the simulated system by adjusting the simulation box volume. This allows simulations to be performed under conditions of constant pressure and temperature (NPT ensemble), mimicking typical laboratory conditions. Different barostats vary in how they couple the volume changes to the pressure difference and whether they allow anisotropic box changes. (UNVERIFIED) barostat algorithm A barostat algorithm that weakly couples the system to an external pressure bath by rescaling the simulation box volume and particle coordinates based on the difference between the current instantaneous pressure and the target pressure. Similar to the Berendsen thermostat, it is computationally simple and effective for equilibration but does not rigorously sample the correct NPT ensemble. It allows isotropic or anisotropic pressure coupling. (UNVERIFIED) barostat=1 pcoupl=Berendsen Berendsen barostat barostat=1 pcoupl=Berendsen A barostat algorithm in which trial changes to the simulation box volume are made at intervals during a molecular dynamics simulation and accepted or rejected with a Monte Carlo criterion. Each trial move rescales the box and the coordinates inside it. It samples the isothermal-isobaric (NPT) ensemble correctly and does not need the virial, the pressure contribution of the forces, to be computed. (UNVERIFIED) barostat=2 Monte Carlo barostat barostat=2 A barostat algorithm in which the simulation box volume is treated as a dynamic variable coupled to a fictitious piston with its own mass, driven by the difference between the internal pressure of the system and the target pressure. This extended Lagrangian approach lets the volume fluctuate around the value that matches the target pressure; on its own it samples the constant-enthalpy, constant-pressure (NPH) ensemble, and combined with a thermostat it samples the NPT ensemble. It changes the volume isotropically, that is equally in all directions. (UNVERIFIED) Andersen barostat A barostat algorithm that extends the Andersen approach by letting all 3 box vectors change, so that both the volume and the shape of the simulation box can fluctuate. The box vectors follow their own equation of motion, driven by the difference between the pressure tensor and the target pressure. This makes it suitable for crystals and other materials under direction-dependent stress, and for phase transitions that change the shape of the box. (UNVERIFIED) pcoupl=Parrinello-Rahman Parrinello-Rahman barostat pcoupl=Parrinello-Rahman A barostat algorithm that treats the box volume as a dynamic variable coupled to a piston, as in the Andersen barostat, and adds a friction term and a random force to the piston's equation of motion, as in Langevin dynamics. The friction and random force damp the oscillations of the box volume that a purely deterministic piston shows. Combined with a thermostat on the particles, it samples the constant-temperature, constant-pressure (NPT) ensemble. (UNVERIFIED) Langevin-piston barostat A Langevin thermostat is a widely used algorithm in molecular dynamics simulations that maintains a constant temperature by mimicking a solvent environment. It modifies Newton's equations of motion by adding two forces to each particle: a frictional drag force that removes kinetic energy, and a random, stochastic force that adds kinetic energy. ntt=3 Langevin thermostat ntt=3 A thermostat algorithm that weakly couples the system to an external heat bath by rescaling particle velocities at each step based on the difference between the current instantaneous temperature and the target temperature. While computationally simple and effective at bringing the system to the target temperature, it does not rigorously generate configurations from the correct canonical (NVT) ensemble. It is often used for equilibration phases rather than production simulations. (UNVERIFIED) ntt=1 tcoupl=berendsen Berendsen thermostat ntt=1 tcoupl=berendsen A thermostat algorithm that maintains temperature by introducing stochastic collisions between randomly selected particles and fictitious heat bath particles. When a particle "collides," its velocity is reassigned from a Maxwell-Boltzmann distribution corresponding to the target temperature. This method correctly generates configurations from the canonical (NVT) ensemble. (UNVERIFIED) ntt=2 tcoupl=andersen Andersen thermostat ntt=2 tcoupl=andersen A deterministic thermostat algorithm that controls temperature by introducing an extended degree of freedom (a thermostat variable) that dynamically couples to the physical system's kinetic energy. This extended variable evolves according to its own equation of motion, causing the physical system's temperature to fluctuate around the target value in a way consistent with the canonical (NVT) ensemble. It avoids the stochastic elements of Langevin or Andersen thermostats. (UNVERIFIED) tcoupl=nose-hoover Nose-Hoover Nosé-Hoover thermostat tcoupl=nose-hoover An algorithm used in molecular dynamics simulations to fix certain geometric parameters, most commonly bond lengths involving light atoms like hydrogen. By constraining these fast vibrations, longer simulation time steps can be used, significantly speeding up the simulation while having minimal impact on slower, large-scale motions. These algorithms iteratively adjust particle positions or use analytical solutions to satisfy the constraint conditions. (UNVERIFIED) constraint algorithm An iterative constraint algorithm widely used in molecular dynamics to fix bond lengths. It adjusts the positions of atoms involved in a constraint at the end of a time step until the constraint condition (e.g., fixed bond distance) is satisfied within a defined tolerance. It handles coupled constraints involving multiple atoms connected together. (UNVERIFIED) SHAKE algorithm A constraint algorithm similar to SHAKE but designed for velocity-Verlet type integrators, ensuring that both positions and velocities satisfy the constraint conditions at the end of each time step. It involves solving constraint equations for positions first, then using those positions to solve for constrained velocities. This maintains consistency between positions and velocities under constraint. (UNVERIFIED) RATTLE algorithm A constraint algorithm that holds bond lengths fixed by correcting the positions after each integration step, solving the constraint equations directly rather than by repeated adjustment. Being non-iterative makes it fast and makes its cost predictable. It is the constraint algorithm GROMACS uses by default. (UNVERIFIED) LINCS linear constraint solver algorithm LINCS algorithm A force calculation algorithm that computes the electrostatic interactions between charged particles in a molecular simulation, including those at long range. Summing every pair directly costs time that grows with the square of the number of particles, so most methods reduce the cost with cutoffs, grids or multipole expansions while keeping acceptable accuracy. Examples are Ewald summation, particle mesh Ewald and reaction field. (UNVERIFIED) electrostatic interaction algorithm A highly efficient grid-based algorithm for calculating long-range electrostatic interactions in periodic systems, based on the Ewald summation technique. It splits the calculation into a short-range part computed directly in real space and a long-range part computed efficiently in reciprocal space using Fast Fourier Transforms (FFTs) on a grid. PME is a standard method for accurate electrostatics in biomolecular simulations. (UNVERIFIED) PME particle mesh Ewald An approximate method for handling long-range electrostatic interactions where direct interactions are calculated only up to a certain cutoff distance. The effect of charges beyond the cutoff is approximated by embedding the cutoff sphere within a continuous dielectric medium representing the solvent. While computationally cheaper than PME, its accuracy depends on the system and parameters chosen. (UNVERIFIED) RF reaction field An efficient algorithm for calculating long-range forces (like electrostatic or gravitational) in particle simulations, conceptually similar to Particle Mesh Ewald (PME). It splits the force calculation into a short-range part computed directly between nearby particles (particle-particle) and a long-range part computed using a grid and Fast Fourier Transforms (particle-mesh). PPPM is widely used in both astrophysics and molecular simulation for its performance and accuracy. (UNVERIFIED) P3M PPPM particle-particle particle-mesh Refers to the classification of forces or potential energy contributions that describe how particles (atoms, ions, coarse-grained beads) affect each other within a simulation's model or force field. These types determine the structure, dynamics, and thermodynamics of the simulated system according to the chosen model. interaction type An interaction between particles where the force or potential energy diminishes slowly with increasing distance, typically decaying as r^−n where n≤3 (e.g., electrostatic interactions decaying as 1/r). Because their influence extends significantly far, these interactions necessitate special computational techniques (like Particle Mesh Ewald (PME) or Ewald sums) in simulations using periodic boundary conditions to accurately account for contributions beyond a simple distance cutoff. They are crucial for the behavior of charged or polar molecules in condensed phases. (UNVERIFIED) long range interaction An interaction between particles where the force or potential energy decreases rapidly with increasing distance, becoming negligible beyond a certain range (typically a few molecular diameters). Van der Waals interactions (often modeled with Lennard-Jones potentials decaying as 1/(r^6) for attraction) are the primary example. These interactions can often be computationally truncated using a cutoff radius in simulations without significant loss of accuracy, improving efficiency. (UNVERIFIED) short range interaction An interaction whose strength and occurrence depend critically on the specific chemical identities, geometries, and relative orientations of the interacting partners. These often involve directional components and chemical complementarity, such as hydrogen bonds, salt bridges (ionic interactions), or π-stacking, playing key roles in molecular recognition, binding specificity, and defined structural motifs. They contrast with more general, less directional interactions like isotropic Van der Waals forces. (UNVERIFIED) specific interaction An interaction energy arising from forces that depend little on the chemical identity or the orientation of the groups involved, chiefly dispersion attraction and excluded-volume repulsion. These forces act between almost any pair of atoms, so they set overall cohesion and packing rather than specific recognition. They contrast with specific, directional interactions such as hydrogen bonds. (UNVERIFIED) non specific interaction energy An interaction between atoms, ions, or molecules that does not involve the sharing of valence electrons to form a covalent bond. These interactions, including Van der Waals forces, electrostatic interactions (encompassing hydrogen bonds, salt bridges, dipole interactions), π-effects, and hydrophobic effects, govern molecular recognition, self-assembly, folding, and the stability of biomolecular complexes. Though individually weaker than covalent bonds (typically 0.1-10 kcal/mol vs >50 kcal/mol), their cumulative effect is critical for biological structure and function. (UNVERIFIED) non-covalent interaction An interaction in which atoms share valence electrons, forming the bonds that hold a molecule together. Typical covalent bonds are much stronger than the thermal energy available at ordinary temperatures, so they rarely break except in chemical reactions. A classical force field represents it as a fixed connection rather than describing the sharing. (UNVERIFIED) covalent interaction A short-range non-covalent interaction between atoms or molecules that combines attraction, mainly from London dispersion and from interactions of permanent or induced dipoles, with a strong repulsion when electron clouds overlap. The dispersion part arises from temporary fluctuations in electron distribution that create short-lived dipoles, and the repulsion arises from the Pauli exclusion principle. It is usually modeled with the Lennard-Jones 12-6 potential, and it determines molecular size, shape and packing and contributes to binding affinity. (UNVERIFIED) Van der Waals interaction A model that represents the effect of the surrounding solvent on a solute in a simulation. It may include individual solvent molecules explicitly, or replace them with a continuous medium carrying averaged properties. The choice sets both the cost of the simulation and which solvent effects can appear in the result. (UNVERIFIED) solvation model A solvation model where individual solvent molecules (e.g., water molecules represented by a specific water model like TIP3P or SPC/E) are explicitly included in the simulation system alongside the solute molecule(s). This approach provides the most detailed, atomistic representation of solute-solvent interactions, capturing specific hydrogen bonding patterns, local solvent structure, and dynamic effects. However, it significantly increases the system size and computational expense due to the large number of solvent particles that must be simulated. (UNVERIFIED) explicit solvent model OBSOLETE. Within the context of solvation models, this typically refers to the use of an explicit solvent model where the solvent is specifically water, represented using a particular force field parameter set (e.g., TIP3P, SPC/E, TIP4P-Ew). This distinguishes it from implicit solvation approaches or explicit simulations using non-aqueous solvents. The term emphasizes that water molecules are being directly simulated using a chosen set of parameters defining their interactions. Retired 2026-08-24. It named a water model but held none: all 18 sit under MOLSIM:000014 water model force field, which its own definition pointed at by naming TIP3P, SPC/E and TIP4P-Ew as the parameter sets involved. The distinction it was reaching for, explicit solvation with water specifically, is already carried by MOLSIM:000066 explicit solvent model, whose definition names the same models. It was also asserted disjoint from MOLSIM:000066, which was wrong, since explicit water is a kind of explicit solvent rather than something separate from it. obsolete water model true A solvation model that approximates the average effect of the solvent environment without explicitly representing individual solvent molecules. Instead, the solvent is treated as a continuous medium characterized by macroscopic properties, primarily its dielectric constant (ϵ) to account for electrostatic screening, and sometimes surface tension to account for nonpolar effects like cavity formation. Common examples include Generalized Born (GB) models and methods based on solving the Poisson-Boltzmann (PB) equation, often combined with a surface area term (SASA); these significantly reduce computational cost compared to explicit solvent. (UNVERIFIED) implicit solvent model A solvation modeling approach that combines elements of both explicit and implicit solvent representations within the same simulation. Typically, a limited number of explicit solvent molecules are retained in a specific region of interest deemed critical (e.g., the active site of an enzyme, the first solvation shell around a solute), while the bulk solvent further away is represented using an implicit continuum model. This strategy aims to capture detailed local solvent effects accurately while maintaining computational efficiency by treating the less critical bulk solvent implicitly. (UNVERIFIED) hybrid solvent model A solvent model designed for compatibility with coarse-grained (CG) solute force fields, where the solvent representation is also simplified to match the reduced resolution of the solute. This can involve explicit CG solvent particles where each particle represents multiple (e.g., 3-4) real water molecules (like MARTINI water), or implicit CG solvent models that adapt continuum concepts to the coarse-grained scale. These models drastically reduce the number of degrees of freedom associated with the solvent, enabling simulations of very large systems or very long timescales. (UNVERIFIED) coarse grained solvent model An emerging class of solvation models where the complex relationship between a solute's structure/properties and its solvation free energy (or related solvent effects) is learned using machine learning (ML) algorithms. These algorithms are trained on large datasets generated from high-fidelity sources like explicit solvent simulations, quantum mechanical calculations, or experimental data. ML solvent models aim to achieve high accuracy approaching that of the training data but at a significantly reduced computational cost, potentially bridging the gap between fast implicit models and expensive explicit models. (UNVERIFIED) machine learning solvent model A solvation model where the electronic structure of the solvent, or its interaction with the solute, is treated using quantum mechanics (QM). This can range from embedding a QM description of the solute within a classical solvent model (either implicit or explicit) using QM/MM techniques, to treating a number of explicit solvent molecules quantum mechanically along with the solute, or using implicit solvent models derived from QM principles (like the SMD model). These models capture electronic polarization, charge transfer, and other quantum effects involving the solvent, providing high accuracy but at a very high computational cost, reserved for specific problems like chemical reactions in solution. (UNVERIFIED) quantum mechanical solvent model A specific type of implicit solvent model that provides an analytical approximation to the electrostatic component of the solvation free energy, based on the Born model for solvation of a spherical ion. It calculates effective Born radii for each atom in the solute, representing the degree of solvent exposure, and uses these radii in a formula that approximates the electrostatic polarization energy typically calculated by solving the more complex Poisson-Boltzmann equation. GB models are computationally much faster than PB methods, making them popular for applications like MD simulations or energy minimization, often paired with a surface area term (GBSA). (UNVERIFIED) GB generalized Born An algorithm used to quantify the degree of resemblance between two objects, often molecules represented by chemical fingerprints or structural features, or between simulation data like conformations or trajectories. These methods employ various mathematical formulas (metrics or coefficients) to compute a numerical score indicating similarity or distance. They are widely used in cheminformatics, drug discovery, and simulation analysis. (UNVERIFIED) similarity calculation algorithm An algorithm that measures how similar 2 molecules are as the size of what they have in common divided by the size of what they cover together. For molecular fingerprints this is the number of features present in both divided by the number present in at least 1 of them. It is the standard measure for fingerprints, runs from 0 for nothing shared to 1 for identical sets, and is also applied to overlapping molecular volumes in shape comparison. (UNVERIFIED) Tanimoto similarity algorithm A similarity calculation algorithm that compares 2 molecules by their overlap while weighting the features unique to each molecule differently, through 2 parameters, α and β. The measure ranges from 0 (nothing shared) to 1 (identical), and the weights can be set to emphasize either the reference (query) molecule or the fit (database) molecule. This asymmetry lets it find molecules that contain the reference as a substructure, or that are contained in it. (UNVERIFIED) General form, for reference: T(A,B) = |A n B| / (alpha*|A - B| + beta*|B - A| + |A n B|), where A and B are the feature sets of the two molecules being compared. alpha weights features unique to A (the reference) and beta weights features unique to B (the query or fit molecule); alpha = beta = 1 reduces to the Tanimoto coefficient, and alpha = beta = 0.5 gives the Dice coefficient. The named variants below differ only in their fixed alpha/beta values, which each of their definitions states. Note this text is a plain-string note: no OWL consumer renders mathematical notation, so it is written to be readable as-is rather than as LaTeX. Requested in GitHub issue #1. Tversky index algorithm A similarity calculation method used to compare 2 sets (e.g., features in molecular fingerprints), closely related to the Tanimoto coefficient and Jaccard index. It measures the overlap between the sets, normalized by the average size of the 2 sets. It is often used in information retrieval and cheminformatics as an alternative similarity metric. (UNVERIFIED) Dice coefficient algorithm A type of statistical property representing the straight-line distance between two points in a multi-dimensional space, which may represent molecules in a property space or conformations by atom coordinates. It is calculated as the square root of the sum of squared differences between corresponding coordinates or features. Lower Euclidean distance implies higher similarity in the context of the chosen space. (UNVERIFIED) Euclidean distance A type of statistical property representing the distance between two points in a multi-dimensional space, calculated as the sum of the absolute differences of their coordinates. The Manhattan distance is an alternative to the Euclidean distance for measuring similarity in clustering and data analysis. This is also known as the city block distance. (UNVERIFIED) Manhattan distance A structural property giving the separation between 2 specified atoms, or between the centres of mass of 2 groups of atoms. It is the basic geometric measure for following interactions such as hydrogen bonds and salt bridges, or the separation between parts of a molecular system, over the course of a simulation. (UNVERIFIED) distance A similarity calculation method specifically used for comparing 2 strings or bit vectors (like binary molecular fingerprints) of equal length. It counts the number of positions at which the corresponding symbols or bits are different. A lower Hamming distance indicates higher similarity between the binary representations. (UNVERIFIED) Hamming distance algorithm A similarity calculation method used for comparing 2 sets, defined as the size of the intersection divided by the size of the union of the sets. It is identical to the Tanimoto coefficient when applied to binary data (like molecular fingerprints). It provides a normalized measure of overlap between feature sets, ranging from 0 (no overlap) to 1 (identical sets). (UNVERIFIED) Jaccard index algorithm A similarity calculation method that measures the overlap between 2 sets, defined as the size of the intersection divided by the size of the smaller of the 2 sets. Unlike Tanimoto/Jaccard, it reaches 1 if one set is a subset of the other. It is useful for assessing containment or sub-structure relationships. (UNVERIFIED) overlap coefficient algorithm A Tversky index algorithm that adds the shape Tversky score and the color (chemical feature) Tversky score of 2 aligned molecules, both weighted towards the fit (database) molecule with α 0.05 and β 0.95. A high score means that the fit molecule is largely contained in the reference (query) molecule in both shape and chemistry, which helps find database molecules that are substructures of the query. The combined score normally ranges from 0 to 2. (UNVERIFIED) fit Tversky combo A specialized variant of the Tversky Index that uses fixed parameters with the alpha (α) value of 0.95, and beta (β) value of 0.05 to heavily weight the reference molecule in the comparison. This asymmetric weighting makes it particularly effective at finding molecules that contain the reference structure as a substructure. RefTversky is commonly used in virtual screening to identify compounds that preserve key structural features of a reference molecule. (UNVERIFIED) ref Tversky A similarity calculation algorithm that combines the scores of 2 or more different similarity measures into a single score. The aim is to capture several aspects of similarity at once, for example the shape of 2 molecules and the chemical features they share. The particular measures chosen and the formula that combines them define each combo score. (UNVERIFIED) combo score algorithm An algorithm designed to find the best path or connection between states or nodes according to some defined criteria, such as minimum cost, shortest distance, or maximum flow. These algorithms are used in various contexts, including finding minimum energy pathways for reactions, constructing minimum spanning trees for clustering, optimizing network routes, or solving complex search problems. They employ different strategies ranging from exact graph traversals to heuristic searches. (UNVERIFIED) optimal transformation path algorithm A greedy algorithm used to find a Minimum Spanning Tree (MST) for a connected, undirected graph with weighted edges. An MST connects all vertices together with the minimum possible total edge weight. Kruskal's algorithm iteratively adds the lowest-weight edge that does not form a cycle until all vertices are connected. Kruskal's algorithm An algorithm that finds the cheapest set of connections joining every node of a network, called a minimum spanning tree, by starting from a single node and repeatedly adding the cheapest link to a node not yet reached. The result is a tree that includes every node with no redundant links. In simulation analysis, such a tree can link similar conformations for clustering. (UNVERIFIED) Prim's algorithm An algorithm for finding the shortest path between a starting node and all other nodes in a graph with non-negative edge weights. It works by iteratively exploring the graph, maintaining the shortest distance found so far to each node, and always extending the path from the node with the current shortest distance. It is fundamental in graph theory and network routing. (UNVERIFIED) Dijkstra's algorithm An informed search algorithm used for finding the shortest path between a starting node and a target node in a graph, commonly used in pathfinding and planning problems. It extends Dijkstra's algorithm by using a heuristic function to estimate the cost from the current node to the target, guiding the search towards the goal more efficiently. Its performance depends on the quality of the heuristic used. (UNVERIFIED) A-star search algorithm A* search algorithm An algorithm that answers a question about a point by looking at whichever recorded points lie closest to it. It is used to classify, to cluster, and to build lists of which atoms are near enough to interact. It makes no assumption about the shape of the data, only that closeness is meaningful. (UNVERIFIED) nearest neighbor algorithm A class of heuristic optimization algorithms inspired by the process of natural selection and evolution. They maintain a population of candidate solutions (chromosomes) that evolve over generations through processes like selection, crossover (recombination), and mutation, guided by a fitness function that evaluates solution quality. Genetic algorithms are often used for complex optimization problems where exact methods are infeasible. (UNVERIFIED) genetic algorithm A probabilistic heuristic algorithm used for finding the global optimum of a function in a large search space, inspired by the annealing process in metallurgy. It starts with an initial solution and iteratively explores neighboring solutions, accepting worse solutions with a probability that decreases over time (controlled by a "temperature" parameter), allowing it to escape local optima. It is a metaheuristic for global optimization. (UNVERIFIED) simulated annealing algorithm A probabilistic heuristic algorithm for finding good paths through graphs, inspired by the foraging behavior of ants who deposit pheromones. Artificial ants construct candidate solutions (paths) by moving through the graph, probabilistically choosing edges based on distance and accumulated artificial pheromone trails, which are updated based on the quality of solutions found. It is often applied to combinatorial optimization problems like the Traveling Salesperson Problem. (UNVERIFIED) ant colony optimization algorithm A class of algorithms designed to solve problems related to flows in networks (graphs where edges have capacities), such as finding the maximum flow possible between a source and a sink node. Examples include the Ford-Fulkerson algorithm and its variations (e.g., Edmonds-Karp). These algorithms are used in logistics, scheduling, network design, and resource allocation problems. (UNVERIFIED) network flow algorithm A binary trajectory format based on the NetCDF standard, used by the Amber simulation package. It efficiently stores atomic coordinates, velocities, forces, and periodic box dimensions from a simulation. This is the modern, recommended trajectory format for Amber. AMBER NetCDF trajectory format AMBER nc format In the vast majority of scientific applications, including molecular dynamics, climate science, and oceanography, a .nc file is a NetCDF file. There is no difference in the format itself. The .nc extension is simply the most common file name suffix for a file written using the NetCDF data model. (UNVERFIED) NetCDF trajectory format Refers to the specific characteristics of a ligand molecule, particularly in the context of its interaction with a receptor molecule (like a protein). This encompasses various aspects, including whether it is bound or unbound, its specific binding pose or conformation, its protonation or tautomeric form, and its electronic state. ligand state The state in which a ligand molecule is physically associated with its receptor molecule, typically residing within a specific binding site or pocket, stabilized by non-covalent interactions. Simulations often aim to characterize the structure, dynamics, and interaction energy of the ligand within the bound state, or calculate the free energy difference between the bound and unbound states (binding free energy). bound state The state in which a ligand molecule is not associated with the receptor molecule, typically solvated freely in the surrounding solvent or located far from the receptor's binding site. Defining the unbound state is necessary for calculating binding free energies, as it serves as the reference state relative to the bound state. It represents the ligand and receptor separated and independently solvated. unbound state In chemical kinetics and simulation, the transition state refers to a specific, high-energy, transient molecular configuration along the reaction coordinate connecting two stable states (e.g., reactant and product, or bound and unbound states). It is connected to the energy barrier that must be overcome for the transition to occur. While challenging to simulate directly due to its short lifetime, characterizing the transition state is crucial for understanding reaction mechanisms and rates, often studied using specialized simulation techniques or QM methods. transition state A state of a system that is locally stable (residing in a minimum on the potential energy surface) but has a higher free energy than the globally stable state. The system can persist in a metastable state for a significant time if the energy barrier to reach a more stable state is high. Examples in biomolecular systems include misfolded protein conformations, transiently bound ligand poses, or specific intermediate states along a functional cycle. metastable state A transient state occurring early in the binding process where the ligand and receptor molecules first come into close proximity and make non-specific contacts, largely driven by diffusion, before potentially rearranging into a specific, stable bound state. The encounter complex represents a collection of loosely associated configurations, often characterized by significant residual motion and non-optimal interactions, preceding the formation of the final bound pose. encounter complex A general state referring to a molecule (either ligand or receptor, or the complex) being surrounded by solvent molecules (typically a water buffer in biological contexts). This term emphasizes the presence and influence of the solvent on the molecule's structure, dynamics, and energetics. Both the bound and unbound states are typically solvated states, and solvation effects are crucial components of binding free energies and conformational preferences. solvated state A specific three-dimensional arrangement or shape adopted by a flexible molecule, such as a protein or ligand. Molecules can typically exist in multiple conformational states, corresponding to different local minima on their potential energy surface, separated by energy barriers. Simulations are often used to explore the accessible conformational states, their relative populations, and the transitions between them, which are essential for molecular function and binding. conformational state The condition of a molecule defined by which of its acidic and basic groups currently carry a proton. It fixes the charge on the molecule and therefore how it interacts with everything around it. It depends on the acidity of the surroundings, so a group buried in a protein may differ from the same group in water. (UNVERIFIED) protonation state One of several forms of a molecule that differ by the position of a hydrogen atom and an accompanying shift of a double bond, and which interconvert readily. The forms are genuinely different molecules for a simulation, because their bonds and charges differ. Choosing the wrong one for a ligand is a common and silent source of error. (UNVERIFIED) tautomeric state An electronic state of a molecule where one or more electrons have been promoted to higher energy levels than in the ground electronic state, typically following the absorption of light (photoexcitation). Excited states have different electron distributions, geometries, and reactivity compared to the ground state and are central to photochemistry and spectroscopy (like fluorescence). Simulating excited states requires quantum mechanical methods (like TD-DFT or CASSCF) as classical force fields are typically parameterized only for the ground electronic state. (UNVERIFIED) excited state A general strategy that specifies how a scientific objective is to be achieved and which procedures are followed to achieve it. In biomolecular science it covers both the physical procedures of an experiment and the calculation protocols of computational work. Knowing which strategy was used tells a reader whether a result came from physical observation or from calculation, and how the work can be repeated. (UNVERIFIED) method An integration algorithm that computes the next position and velocity from the current ones and the current force alone. It is the simplest possible scheme and the easiest to reason about. It is not used for ordinary molecular dynamics, because the energy of a system integrated this way drifts steadily upward, although Euler-type schemes are used for Brownian dynamics. (UNVERIFIED) Euler integrator OBSOLETE. A type of statistical property that measures the strength and direction of the linear association between two variables. It is a value between -1 and 1, where 1 indicates a perfect positive correlation and -1 indicates a perfect negative correlation. The correlation coefficient is widely used to identify relationships between different properties in a simulation. This is often reported as r or rho. Obsoleted 2026-07-29: duplicated STATO:0000142. STATO is the canonical OBO home for statistics; reused in place of this term. Its four children were re-parented under STATO:0000142 in the same edit - retiring a parent necessarily re-points every child that named it. obsolete correlation coefficient true OBSOLETE. Pearson's correlation coefficient (often denoted as 'r') measures the strength and direction of the linear relationship between two continuous variables. It assumes the data are approximately normally distributed and calculates the covariance of the two variables divided by the product of their standard deviations. It is perhaps the most widely used measure of linear correlation. Obsoleted 2026-07-29: duplicated STATO:0000280. STATO is the canonical OBO home for statistics; reused in place of this term. obsolete Pearson's correlation coefficient true A specific type of `orientation` descriptor that represents the cosine of the rotation of a vector orthogonal to a given axis. tilt orientation HOLE is a computer program used to analyze the dimensions and shape of internal pathways or tunnels within a biomolecular structure. It operates by computationally squeezing a flexible sphere along a channel axis to determine the maximum pore radius at every point without overlapping the surrounding atoms. https://doi.org/10.1016/S0263-7855(97)00009-X HOLE MOLE is a software algorithm designed to automatically locate and characterize channels, tunnels, and pores within protein structures. It utilizes a Voronoi diagram approach to efficiently map the void space inside a molecule and identify paths connecting internal cavities to the surface. MOLE A resampling method that estimates the uncertainty of a computed quantity by drawing many new samples of the same size from the existing data, with repetition allowed, and recomputing the quantity from each. The spread of the recomputed values estimates the standard error or confidence interval of the original value. It makes no assumption about how the data are distributed, but it treats data points as independent, so correlated data such as consecutive simulation frames are usually resampled in blocks. (UNVERIFIED) bootstrap analysis It is a kinetic network model used to analyze complex conformational dynamics, such as protein folding, ligand binding, and structural rearrangements, by stitching together many short molecular dynamics (MD) simulations into a global long-timescale picture. It divides the conformations visited in simulations into discrete states and counts the transitions between them, assuming the next state depends only on the current one and not on the path taken there. MSM Markov state model analysis A statistical analysis that measures how different 2 probability distributions are by averaging the relative entropy (Kullback-Leibler divergence) of each from their mean distribution. The result is symmetric, giving the same value whichever distribution is taken first, and bounded, lying between 0 and 1 when base-2 logarithms are used; plain relative entropy has neither property. It is used to compare the conformations sampled by 2 simulations. (UNVERIFIED) Jensen-Shannon divergence analysis A statistical measure of how much information is lost when one probability distribution is used in place of another. It is not symmetric: the value depends on which distribution is treated as the reference. It is 0 only when the 2 are identical and has no upper limit, which is why the symmetric bounded version is often preferred. (UNVERIFIED) Kullback-Leibler divergence A statistical method that re-expresses data with many variables in terms of a few, chosen to keep as much of the structure as possible. A protein trajectory has 3 coordinates per atom, which cannot be looked at directly. Reducing it to 2 or 3 numbers per frame is what makes it possible to see where the system went. (UNVERIFIED) dimensionality reduction technique analysis A statistical method that finds the directions along which data vary most, and re-expresses the data in terms of the few largest of them. Applied to a trajectory, the largest directions are the collective motions that account for most of the movement. A protein with thousands of coordinates can then be described by 2 or 3 numbers per frame. (UNVERIFIED) PCA principal component analysis A data analysis technique is a specific computational method used to extract meaningful scientific insights from the raw data generated by a simulation. These techniques can range from simple calculations, like measuring the distance between two atoms over time, to complex statistical methods that identify large-scale molecular motions. Applying these techniques is the crucial step that transforms a trajectory of atomic coordinates into an understandable story about the molecule's behavior. data analysis technique Clustering analysis encompasses algorithms designed to partition a dataset into distinct groups or clusters, such that data points within the same cluster are more similar to each other than to those in other clusters based on chosen features. In molecular simulation, clustering is frequently applied to trajectory data to identify distinct conformational states visited by the system. Various algorithms exist, differing in how similarity and clusters are defined (e.g., k-means, hierarchical clustering). clustering analysis A hypothesis testing analysis that decides whether to reject a null hypothesis by computing a test statistic from the sample data and comparing it with a critical value, or by computing a p-value. A p-value is the probability of observing data at least as extreme as those seen if the null hypothesis were true. Examples include t-tests, chi-squared tests and analysis of variance (ANOVA). (UNVERIFIED) statistical test analysis A statistical test that judges whether 2 sets of values come from the same distribution, or whether 1 set comes from a given reference distribution, by measuring the largest gap between their cumulative distributions. It makes no assumption about the shape of the distributions, but it does assume that the values are independent, which consecutive simulation frames are not. It is used, for example, to check whether 2 simulations sampled the same distribution of a property. (UNVERIFIED) Kolmogorov-Smirnov test analysis A statistical analysis that computes, from sample data such as simulation results, a range of values that would contain the true value of a quantity in a chosen share of repeated experiments, such as 95%. It states how uncertain an estimate is when the data are limited or noisy. A wider interval means greater uncertainty. (UNVERIFIED) confidence intervals analysis A statistical method that measures how far predicted values lie from observed ones. Which measure to use depends on what a large error should cost: squaring the differences makes outliers dominate, taking their size does not. Reporting more than one is usual, because each hides something the others show. (UNVERIFIED) error metrics analysis A statistical measure giving how far a computed average is likely to lie from the true average, based on how much the data scatter and how many independent samples there are. It shrinks as more independent data is added, unlike the spread of the data itself, which does not. For a trajectory the count of independent samples is far smaller than the number of frames, so it is usually estimated with methods that account for this, such as block averaging. (UNVERIFIED) SEM standard error of the mean analysis An error metric giving the typical size of the difference between predicted and observed values, computed by squaring the differences, averaging them and taking the square root. Squaring first makes large errors count disproportionately, so a single bad prediction moves it noticeably. It is in the same units as the quantity itself, which makes it directly interpretable. (UNVERIFIED) RMSE root mean square error analysis An error metric giving the average size of the difference between predicted and observed values, ignoring sign. Because it does not square the differences, every error counts in proportion to its size and one large error does not dominate. It is the measure to prefer when outliers are expected and should not be allowed to control the score. (UNVERIFIED) MAE mean absolute error analyisis An error metric giving the average of the squared differences between predicted and observed values. It is the quantity most fitting procedures minimise, because squaring makes it smooth and easy to differentiate. Its units are the square of the quantity's, which is why the square root is usually reported instead. (UNVERIFIED) MSE mean squared error analysis A widely used all-atom force field parameter set from the CHARMM family, optimized for proteins, lipids, nucleic acids, and carbohydrates. Developed by the MacKerell Lab at the University of Maryland, it provides the mathematical parameters (atomic charges, masses, bond lengths, and angles) necessary to calculate the physical forces dictating the structure and dynamics of biomolecules. It represents a significant update over earlier CHARMM versions (like CHARMM22/27), incorporating improved parameters derived from high-level quantum calculations and experimental data. The 'm' variant (CHARMM36m) includes further refinements, particularly for proteins. CHARMM36 A statistical method that measures how well a model predicts data it was not fitted to, by repeatedly holding part of the data back, fitting on the rest, and testing on the held-back part. A model can always be made to fit the data it was built from, so testing on that data proves nothing. This is the standard guard against that mistake. (UNVERIFIED) cross-validation errors analysis A validation method that puts an uncertainty on a result by resampling the data it came from and recomputing it many times. It is used to show that a conclusion does not rest on a few particular data points. A large spread across the resamples shows that the result is not reliable. (UNVERIFIED) BSE bootstrapped error estimates analysis A data analysis method that judges whether a simulation has sampled enough for its results to be stable, and puts an error estimate on those results. It works from time series of the quantity of interest, or by resampling the data statistically. Without both parts, a reported value has no stated reliability. (UNVERIFIED) convergence and uncertainty quantification analysis A statistical method that estimates the uncertainty of an average taken from a trajectory, by dividing the data into blocks and looking at how the block averages scatter. It is needed because successive frames of a trajectory resemble each other, so the usual formula for the uncertainty of a mean underestimates it badly. Enlarging the blocks until the estimate stops changing gives the honest figure. (UNVERIFIED) block averaging analysis A data analysis method that compares results obtained while a parameter or collective variable is changed in one direction with those obtained while it is changed back, and reports the difference between them. A large difference means the system did not stay near equilibrium, or that sampling was insufficient. Showing that the difference is small is part of demonstrating that a free energy calculation has converged. (UNVERIFIED) hysteresis analysis A data analysis method that computes quantitative measures of whether a simulation has sampled enough for its results to have stopped changing. Measures include the stability of a running average, the size of fluctuations in a key property, and the number of transitions observed between states. Which measure is appropriate depends on the kind of simulation and on what is being calculated. (UNVERIFIED) convergence metrics analysis A data analysis method that measures how far the configurations sampled in neighbouring simulation windows or thermodynamic states coincide. Enough overlap is needed to combine the windows with reweighting methods such as the weighted histogram analysis method (WHAM) or the multistate Bennett acceptance ratio (MBAR), and too little makes the combined free energy unreliable. Poor overlap shows that more windows or longer sampling are needed. (UNVERIFIED) overlap metrics analysis A method that studies a scientific problem by calculation on a computer, for example by simulating molecular behavior, building molecular models or analyzing the resulting data. It is distinct from an experimental method, which observes a physical sample, and from a theoretical approach, which supplies the physical principles a calculation is built on. It is defined by its algorithms and by the settings given to the program that runs it. (UNVERIFIED) computational method A method consisting of the physical principles and mathematical framework a calculation is built on, such as quantum mechanics or classical mechanics with an empirical potential. It sits above the algorithm and the software, and fixes what the calculation is able to represent at all. Two methods sharing an approach can still differ in every other respect. (UNVERIFIED) theoretical approach A statistical metric that scores how well a model ranks a set of compounds, by checking for each pair whether the predicted order matches the measured order and weighting the pair by the difference between their measured values. A wrong order for 2 compounds with very different measured values therefore counts more than a wrong order for 2 similar ones. It is used where a model is meant to guide a choice between compounds rather than to reproduce a number. (UNVERIFIED) predictive index analysis A criterion applied when choosing which pairs of compounds to compare in a relative binding free energy study. The choice strongly affects whether the calculation converges, because a pair that differs by 1 atom is far easier to transform than a pair that differs by a whole ring. Such criteria cover how similar the compounds are, the range of their measured affinities, the quality of the available data and how practical the calculation is. (UNVERIFIED) ligand pairs selection criteria A ligand selection criterion favouring pairs of molecules that share a common core and differ only by small changes. Such pairs can be transformed into one another alchemically with few atoms changing, which helps the calculation converge and lets errors cancel. Similarity of this kind is often judged by eye or with molecular fingerprint measures. (UNVERIFIED) structural similarity A ligand selection criterion favouring molecules that resemble each other in chemical character, such as their functional groups, polarity, charge and hydrogen bonding, often measured with chemical fingerprints or descriptors. It is related to structural similarity but looks at the chemical nature of the difference rather than only at a shared core. Choosing chemically similar pairs helps keep an alchemical transformation smooth. (UNVERIFIED) chemical similarity A ligand selection criterion requiring the measured binding strengths of the chosen compounds to fall within a stated range, or to differ by no more than a stated amount. Differences smaller than the typical error of the calculation cannot be told apart, and the measured values must lie within what the assay can measure reliably for the comparison to mean anything. (UNVERIFIED) binding affinity range A ligand selection criterion requiring the chosen set as a whole to span a range of structures and chemistries, even though individual pairs within it should be similar. A set of near-identical compounds tests a method over a narrow region and says nothing about its general reliability. The 2 requirements apply at different levels: similarity within a pair, diversity across the set. (UNVERIFIED) structural and chemical diversity This practical criterion emphasizes selecting ligands for which reliable experimental data, particularly binding affinity measurements (like Ki, Kd, or IC50), already exist. Having high-quality experimental data is essential for validating the computational predictions and assessing the accuracy of the methods used. Availability of structural data (e.g., co-crystal structures) is also highly valuable. experimental data availability A ligand selection criterion based on computed or measured properties of the compounds, such as molecular weight, lipophilicity, charge and the number of rotatable bonds. These properties predict how difficult a compound will be to simulate as well as how drug-like it is. A charged or very flexible compound needs far more sampling than a neutral rigid one. (UNVERIFIED) chemical properties A ligand selection criterion based on how large and of what kind the difference between 2 candidate compounds is. Small, local changes, such as adding a methyl group or swapping a halogen for another, usually converge quickly, while opening a ring or changing the net charge does not. Smaller changes generally give more reliable relative free energy results. (UNVERIFIED) structural changes structural change A ligand selection criterion asking whether the calculations a candidate would require can be completed with the time and computing available. A charged or highly flexible ligand may need many times the sampling of a neutral rigid one. It is applied before the calculation, since discovering the answer afterwards wastes the whole effort. (UNVERIFIED) computational feasibility A ligand selection criterion favouring molecules and modifications that bear on an ongoing drug discovery project or a stated therapeutic aim. It is a judgement about purpose rather than about chemistry, so it cannot be computed. It is recorded because it explains why a particular set of compounds was chosen. (UNVERIFIED) relevance to drug discovery A ligand selection criterion based on how good the experimental structure of the complex is, judged by resolution, how well the model fits the data, and whether the ligand is clearly resolved. A poorly resolved ligand may be placed wrongly, and a calculation started from it inherits that error. It is assessed before the structure is used, not after a result looks wrong. (UNVERIFIED) structural data quality A ligand selection criterion based on whether the compound stays bound and in a consistent pose during a trial simulation. A ligand that drifts out of its site, or flips between poses, gives a free energy that depends on how long the run was. Testing it with a short simulation before committing to a full calculation is cheaper than discovering it afterwards. (UNVERIFIED) stability in simulations A ligand selection criterion based on how tightly a compound binds relative to its size, usually measured as binding free energy divided by the number of heavy, that is non-hydrogen, atoms. It is used because binding strength tends to rise with size, so a large weak binder can look better than a small efficient one. Compounds that bind efficiently for their size are often preferred as starting points in early drug discovery. (UNVERIFIED) ligand efficiency The GROMACS mdp format is the specific simulation parameter format used by the GROMACS software package. The file, which typically has a ".mdp" extension, contains a series of keyword-value pairs that control every aspect of the simulation run, from the integrator and time step to the thermostat and pressure coupling methods. For GROMACS users, creating and editing the mdp file is the primary way to define the simulation protocol. GROMACS mdp format A reusable software platform that provides a common structure, tools and libraries on which simulation workflows or coupled simulation programs are built, at a higher level than a single tool or script. It may standardise how simulations are set up, how data are managed, or how separate programs exchange data during a run. Examples include workflows that automate free energy calculations and interfaces that couple a quantum chemistry program to a molecular dynamics program. (UNVERIFIED) framework simulation framework A software program that builds or runs a computational model of a physical system, such as a simulation engine or a modelling and setup tool. What separates it from other software is that its subject is a model of something real. (UNVERIFIED) simulation and modeling software A text-based file format from the Amber software that stores the movement of atoms over time from a simulation. It contains a series of snapshots, or frames, where each frame lists the x, y, and z coordinates of every atom at a specific time point. It may also store the x, y, and z simulation box sizes. AMBER mdcrd format mdcrd format Application programs specifically designed to perform molecular dynamics (MD) simulations, which calculate the motion of atoms and molecules over time based on classical mechanics and a chosen force field. These packages typically include capabilities for system preparation, simulation execution (often optimized for high-performance computing), and basic trajectory analysis. Examples include AMBER, GROMACS, NAMD, CHARMM, Desmond, and LAMMPS. molecular dynamics engine A software program that solves for the arrangement of electrons in atoms and molecules using quantum mechanics. It implements methods such as Hartree-Fock, density functional theory and the correlated methods built on them. It is what supplies an energy and a set of forces when a force field is not accurate enough. (UNVERIFIED) quantum chemistry engine A computational tool used to predict the preferred binding orientation and affinity (scoring) of one molecule (typically a small ligand) to another (usually a larger receptor like a protein or nucleic acid). Docking tools employ search algorithms to explore possible binding poses within a target site and use scoring functions to estimate the binding strength for each pose. Widely used in drug discovery and understanding molecular recognition, examples include AutoDock, Vina, Glide, GOLD, and DOCK. docking tool A software program that turns data into images, either as interactive 3-dimensional views of molecular structures, surfaces, densities and trajectories, or as plots and graphs of numerical results. It lets a person look at a system or a result rather than only measure it. Many faults in a prepared system, such as a molecule sitting half outside its box, are obvious on sight and invisible to any numerical check. (UNVERIFIED) visualization tool Software designed to process raw data generated from biomolecular simulations (like trajectories, energy files) to calculate specific quantitative properties and extract meaningful insights. Analysis tools implement algorithms for calculating metrics like RMSD, RMSF, radius of gyration, hydrogen bonds, interaction energies, principal components (PCA), free energy estimations (MM/PBSA), clustering, etc. Examples include cpptraj (AmberTools), GROMACS analysis utilities, MDAnalysis (Python library), and specialized scripts or programs. analysis tool A tool focused on manipulating, converting, filtering, or extracting information from data files. This can involve tasks like converting file formats, extracting specific columns or frames, merging datasets, calculating simple statistics, or preparing data for input into other analysis or visualization tools. data processing tool A specific type of data processing tool dedicated to converting molecular structure, trajectory, or data files from one format to another (e.g., PDB to MOL2, XTC to netcdf). file conversion tool A script specifically designed to automate a series of steps in a workflow that would otherwise need to be performed manually. In simulations, this could involve automatically setting up multiple simulation replicas, submitting jobs to a computing cluster, checking job status, and organizing output files, thereby reducing manual effort and potential errors. Workflow management systems often rely heavily on automation scripts. (UNVERIFIED) automation script A script written to perform specific analysis tasks on data generated from simulations, often going beyond the standard capabilities of packaged analysis tools or requiring custom calculations. Examples include scripts to calculate specific geometric parameters over time, parse log files for specific energy terms, perform advanced statistical analysis or time-series analysis on trajectory data, or generate data for custom plots. data analysis script A script that defines and executes an entire sequence of computational tasks constituting a specific workflow, often involving multiple steps like data preparation, running simulations (potentially submitting to clusters), performing analysis, and summarizing results. These scripts orchestrate the execution of various tools and other scripts, manage intermediate data, and handle dependencies between steps, ensuring the entire process runs smoothly and reproducibly. Tools like Snakemake or Nextflow help in creating robust workflow scripts. (UNVERIFIED) workflow script A reusable part of a larger software system that provides a defined set of functions to whatever calls it. Unlike a program, it is not run on its own; it is built into something else. A library that reads a file format is the usual example. (UNVERIFIED) software component A software program that helps in writing, building, installing, testing or maintaining other software, rather than in doing any science itself. Examples are compilers, build systems, package managers and version control systems. A simulation result can depend on which versions of these tools produced the software that computed it. (UNVERIFIED) software development and management tool A software program that produces the topology and force field parameters a simulation needs for a molecule: which atoms are bonded, what atom type and partial charge each atom has, and the parameters for each bond, angle and dihedral. Some such programs assign existing parameters by atom type, while others derive new charges, fit new parameters or build residue templates. A wrong assignment gives a system that runs normally but represents the wrong molecule. (UNVERIFIED) topology and parameterization tool A physical theory describing how matter behaves at the scale of atoms and subatomic particles, where energy and angular momentum take discrete values. It is what chemical bonding and molecular electronic structure are explained by. In this field its equations are solved approximately to predict molecular properties. (UNVERIFIED) QM quantum mechanics A minimal basis set in which each Slater-type atomic orbital is approximated by a fixed combination of 2 Gaussian functions. Gaussian functions are used because the integrals they need are much faster to compute than those of Slater-type orbitals. With only 2 the approximation is poor, so it is used to demonstrate the method rather than to obtain results. (UNVERIFIED) STO-2G A commercial quantum chemistry engine for electronic structure calculations with ab initio, density functional and semi-empirical methods. It can also combine quantum and classical models of different parts of a system with its ONIOM method, and is often used to compute the reference data from which force field parameters, such as atomic charges, are derived. It is not designed to be a classical molecular dynamics engine for large biomolecules. (UNVERIFIED) Gaussian A quantum chemistry engine providing a broad range of wavefunction-based and density functional methods. The name is shared by 2 separately maintained programs, developed in the United States and the United Kingdom, and their inputs are not interchangeable. Which one produced a result therefore has to be recorded. (UNVERIFIED) GAMESS An open-source quantum chemistry engine specializing in high-accuracy ab initio methods, such as coupled cluster theory and symmetry-adapted perturbation theory, alongside Hartree-Fock and density functional theory. It is written in C++ with a Python interface, so calculations can be set up and combined in Python scripts. Symmetry-adapted perturbation theory splits the interaction between 2 molecules into physical parts such as electrostatics and dispersion. (UNVERIFIED) Psi4 A quantum chemistry engine, free for academic use, that offers a wide range of methods, including density functional theory, Hartree-Fock, semi-empirical methods, Møller-Plesset perturbation theory, coupled cluster and multi-reference methods. It is known for efficient calculations on large molecules, especially with density functional theory, and is widely used for spectroscopic properties and for QM/MM calculations on biomolecules. (UNVERIFIED) Orca A commercial quantum chemistry engine offering Hartree-Fock, density functional theory, coupled cluster theory, methods for excited electronic states and QM/MM calculations, which treat a small region with quantum mechanics and the rest classically. Its primary focus is the electronic structure of molecules. It is not designed for large-scale classical molecular dynamics simulations. (UNVERIFIED) Q-Chem An integrated suite of open-source programs for electronic structure calculations and materials modeling based on density functional theory, plane waves and pseudopotentials. It is designed for periodic systems such as crystals and surfaces, and can run ab initio molecular dynamics in which the forces come from density functional theory at every step. It is used mainly in condensed matter physics and materials science rather than for large solvated biomolecules. (UNVERIFIED) Quantum ESPRESSO A versatile open-source computational chemistry software package developed by Pacific Northwest National Laboratory (PNNL), designed for high-performance computing. It offers a broad range of capabilities, including classical MD simulation, various quantum mechanical methods (DFT, Hartree-Fock, coupled cluster), and hybrid QM/MM methods. Its ability to handle both MD and QM makes it suitable for complex biomolecular simulations requiring multi-scale approaches on large parallel computers. (UNVERIFIED) NWChem A quantum chemistry engine for highly accurate ab initio calculations, best known for its multi-reference methods, such as CASSCF, CASPT2 and MRCI, and its coupled cluster methods, such as CCSD(T), alongside Hartree-Fock and density functional theory. Multi-reference methods describe molecules whose electrons cannot be represented by a single arrangement, such as many transition metal complexes and molecules in excited states. It is used to compute accurate energies, properties and potential energy surfaces, often as benchmarks for cheaper methods. (UNVERIFIED) Molpro An open-source quantum chemistry engine, developed at the University of Washington, that focuses on time-dependent and relativistic electronic structure methods and on their combination. It treats time dependence, relativistic effects, electron correlation, electron-nuclear coupling and spin consistently, and can include external magnetic fields. It serves both as a platform for developing new methods and as a program for running such calculations. (UNVERIFIED) Chronus Quantum A Python library for simulating the dynamics of open quantum systems, which exchange energy with their surroundings rather than being isolated. Its subject is quantum optics and quantum information rather than molecular electronic structure. (UNVERIFIED) QuTiP A non-covalent interaction occurring between aromatic rings, arising from a combination of electrostatic interactions between their quadrupole moments (electron-rich faces vs electron-poorer edges) and Van der Waals (dispersion) forces. The preferred geometry (e.g., stacked, parallel-displaced, T-shaped) depends on the specific aromatic systems and influences the interaction strength. They are important in protein structure (stacking of Phe, Tyr, Trp, His residues) and the stability of DNA/RNA helices (base stacking). (UNVERIFIED) π-π interactions π-π interaction The observed tendency for nonpolar molecules or molecular groups to associate with each other when in an aqueous (water) environment, effectively minimizing their disruptive contact with water molecules. This phenomenon is primarily entropically driven by the favorable increase in the entropy of water molecules that are released from ordered "cages" (solvation shells) around the nonpolar surfaces when they aggregate. It is a major driving force for protein folding (burying hydrophobic residues), membrane formation, and ligand binding. hydrophobic interactions hydrophobic interaction An interaction between charged or polar groups arising from the attraction and repulsion of their charges. It reaches much further than any other non-bonded interaction: for 2 ions the energy falls off only as the inverse of the distance. That long reach is why a periodic simulation needs a special method to compute it rather than simply cutting it off. (UNVERIFIED) electrostatic interactions electrostatic interaction An electrostatic interaction in which a cation is attracted to the electron-rich face of an aromatic ring. The attraction comes from the negative electrostatic potential above and below the ring plane, which arises from the ring's quadrupole moment. In proteins, typical partners are a metal ion or the side chain of lysine or arginine with the ring of phenylalanine, tyrosine or tryptophan, and these interactions contribute to structure stabilization and ligand binding. (UNVERIFIED) cation-π interactions cation-π interaction A specific, directional non-covalent interaction formed between a hydrogen atom covalently bonded to a highly electronegative atom (the donor, D, typically N, O, or F) and another nearby electronegative atom (the acceptor, A, also typically N, O, or F) which possesses a lone pair of electrons. Represented as D-H···A, it has characteristics of both electrostatic attraction (between the partially positive H and partially negative A) and weak orbital overlap/covalency. hydrogen bond Hydrogen bonds formed specifically between the atoms of the polypeptide backbone in proteins or peptides. The most common form involves the amide hydrogen (N-H group) of one residue acting as the donor and the carbonyl oxygen (C=O group) of another residue (often i→i+4 in helices or between strands in sheets) acting as the acceptor. backbone hydrogen bonds backbone hydrogen bond Hydrogen bonds where at least one of the participating atoms (either the donor hydrogen attached to its electronegative atom, or the acceptor electronegative atom) belongs to an amino acid side chain, rather than exclusively involving backbone atoms. These interactions can occur between two side chains, or between a side chain and the backbone, contributing to the stability of the protein's tertiary and quaternary structure, protein-ligand interactions, and enzyme active site chemistry. side-chain hydrogen bonds side-chain hydrogen bond Hydrogen bonds formed specifically between an atom on the polypeptide backbone (either the amide N-H donor or the carbonyl C=O acceptor) and an atom within an amino acid side chain (acting as acceptor or donor, respectively). These interactions directly link the regular backbone structure to the specific side chain identities and conformations, playing crucial roles in stabilizing protein folds (tertiary structure) and positioning side chains for function. backbone-side chain hydrogen bonds backbone-side chain hydrogen bond A covalent bond joining the carbonyl carbon of one amino acid to the nitrogen of the next, formed with the loss of a water molecule. Its partial double-bond character holds the 6 atoms around it flat and makes rotation about it slow. That rigidity is why a protein backbone has only 2 freely rotating angles per residue. (UNVERIFIED) peptidic bond A specific type of covalent bond formed by the oxidation of the thiol (-SH) groups of two cysteine amino acid residues, creating a sulfur-sulfur bond (-S-S-). These bonds can form between cysteines within the same polypeptide chain (intramolecular) or between different chains (intermolecular), acting as strong cross-links that significantly stabilize the folded tertiary or quaternary structure of many proteins, particularly extracellular ones. Formation and breakage of disulfide bridges are important biochemical processes. (UNVERIFIED) disulfide bridges disulfide bridge A theoretical statistical mechanics construct representing a probability distribution for the state of a system, defined by a specific set of macroscopic variables held constant (e.g., NVT, NPT). In molecular simulation, it dictates the thermodynamic conditions and integration rules applied to generate a trajectory. (UNVERIFIED) ensemble statistical ensemble thermodynamic ensemble A statistical ensemble representing systems where the number of particles (N), the system volume (V), and the temperature (T) are held constant; hence often called the NVT ensemble. The system is considered to be in thermal contact with an infinitely large heat bath at temperature T, allowing energy exchange. Simulations performed in the NVT ensemble generate configurations according to the Boltzmann probability distribution appropriate for that temperature, making it suitable for studying equilibrium properties under constant temperature and volume conditions. (UNVERIFIED) NVT ensemble canonical ensemble The isothermal-isobaric ensemble (often called the NPT ensemble) is a statistical mechanical framework for a system with a fixed number of particles (N), constant temperature (T), and constant pressure (P). Unlike canonical ensembles, volume (V) is allowed to fluctuate, making it ideal for modeling real-world condensed-phase laboratory conditions. NPT ensemble isothermal-isobaric ensemble A statistical ensemble representing completely isolated systems, where the number of particles (N), the system volume (V), and the total energy (E) are all held constant; hence often called the NVE ensemble. There is no exchange of energy or particles with the surroundings. Simulations performed in the NVE ensemble directly follow Newtonian laws of motion (integrating Hamilton's equations), and conservation of total energy serves as a key check of simulation stability and accuracy, but it's less commonly used for equilibrium sampling compared to NVT or NPT as temperature can fluctuate. (UNVERIFIED) NVE ensemble microcanonical ensemble A statistical ensemble representing open systems where the chemical potential (μ), the system volume (V), and the temperature (T) are held constant; hence often called the μVT ensemble. The system can exchange both energy (with a heat bath at T) and particles (with a particle reservoir at chemical potential μ). Simulations in this ensemble are useful for studying systems where the number of particles can change, such as adsorption phenomena, phase equilibria (vapor-liquid), or systems involving chemical reactions where concentrations are fixed rather than particle numbers. (UNVERIFIED) µVT ensemble grand canonical ensemble Protein structure determination methods refer to the experimental techniques used to obtain high-resolution three-dimensional atomic coordinates of protein molecules. These methods provide the foundational structural data used in biomolecular simulations, structure-based drug design, and understanding protein function. The primary techniques include X-ray crystallography, NMR spectroscopy, and cryo-electron microscopy. protein structure determination method A protein structure determination method that works out the 3-dimensional structure of a molecule from the pattern produced when X-rays are diffracted by a crystal of that molecule. The diffraction pattern gives the arrangement of electrons as an electron density map, into which the atomic structure is built. It can give very high-resolution structures. (UNVERIFIED) X-ray X-ray crystallography Nuclear Magnetic Resonance (NMR) spectroscopy is an experimental technique used to determine protein structures, typically in solution, by exploiting the magnetic properties of atomic nuclei. It measures distances between specific pairs of atoms (primarily hydrogens using the Nuclear Overhauser Effect) and torsion angle restraints based on spin-spin coupling. These experimentally derived restraints are then used to calculate an ensemble of structures consistent with the data. NMR nuclear magnetic resonance A protein structure determination method that images molecules and complexes, rapidly frozen in a thin layer of non-crystalline (vitreous) ice, with an electron microscope. Image processing averages the images of thousands to millions of individual particles to reconstruct a 3-dimensional density map, often at high resolution. It is especially useful for large, flexible complexes that are difficult to crystallize. (UNVERIFIED) Cryo-EM cryogenic electron microscopy A method or protocol for assigning partial charges to the atoms of a small molecule, such as a ligand, for use in a classical force field. Electrostatic interactions strongly affect how a ligand binds, so the charges must be accurate and derived in a way consistent with the force field. Common models, such as RESP and AM1-BCC, fit charges to the electrostatic potential from a quantum calculation or correct cheaper charges with fitted terms. (UNVERIFIED) ligand charge model A widely used method for assigning atom-centered partial charges to organic molecules, particularly suitable for general force fields like GAFF (stands for Austin Model 1 - Bond Charge Correction). It involves a two-step process: first, charges are calculated using the computationally inexpensive semi-empirical AM1 quantum mechanical method; second, empirical Bond Charge Corrections (BCCs) are added to these initial charges based on atom and bond types to make them better reproduce the electrostatic potential obtained from higher-level QM calculations (typically HF/6-31G*). AM1-BCC aims to provide a balance between speed and accuracy for charging diverse organic ligands. (UNVERIFIED) AM1-BCC A popular method for deriving high-quality partial atomic charges for molecules, frequently used in conjunction with the AMBER force fields. It involves calculating the quantum mechanical electrostatic potential (ESP) on a grid of points around the molecule (typically using Hartree-Fock or DFT methods), and then fitting atom-centered point charges to reproduce this ESP using a least-squares algorithm. Crucially, 'restraints' (typically hyperbolic penalties) are applied during the fitting process to prevent unrealistically large charges on buried atoms and ensure transferability, often performed in multiple stages for chemical equivalence. (UNVERIFIED) RESP restrained electrostatic potential A script written by a researcher for one specific purpose in their own project, rather than distributed as part of a tool. It exists because no available program did quite what was needed. It is worth recording because a result may depend on it entirely, and it is usually the least documented part of a workflow. (UNVERIFIED) custom script A step within the preparation or equilibration phase where the system's temperature is gradually raised or adjusted to the desired target temperature for the production simulation. This is typically done by coupling the system to a thermostat while velocities are assigned or rescaled. It brings the velocities to the distribution that matches the target temperature before equilibration or production runs. (UNVERIFIED) heating thermalization A crucial phase in simulation preparation, following initial setup and minimization/thermalization, where the system is simulated under the target ensemble conditions (e.g., NVT or NPT) for a period to allow it to relax and reach a stable state. Key properties like temperature, pressure, density, potential energy, and RMSD are monitored to ensure the system is well-equilibrated before starting the production simulation intended for data collection. This phase allows the system to 'forget' its initial, potentially artificial starting configuration. equilibration A specific simulation execution technique used to enhance sampling along a chosen reaction coordinate or order parameter, often for calculating the potential of mean force (PMF) or free energy profile associated with a process like ligand unbinding or a conformational change. It involves running multiple simulations, each confined ('biased') by a potential (the 'umbrella') to sample a specific window along the coordinate, with subsequent analysis required to unbias the results and reconstruct the overall profile. It falls under the category of enhanced sampling methods. US umbrella sampling A stage of a simulation run in which data is collected, after the system has been prepared and allowed to settle. It is run under the conditions the study is about, for long enough that averages taken from it stop drifting. Only this part of a trajectory is normally analysed; the settling period before it is discarded. (UNVERIFIED) production A simulation execution designed to compute the free energy difference between 2 states, such as a ligand bound and unbound, or the free energy profile along a reaction coordinate. Common methods are free energy perturbation, thermodynamic integration and umbrella sampling. These calculations are computationally demanding but are central to estimating binding affinities, reaction barriers and relative stabilities. (UNVERIFIED) free energy calculation An enhanced sampling simulation technique that enhances traditional steered molecular dynamics (SMD) to calculate binding affinities, mechanical unfolding pathways, and free energy profiles (e.g., Potential of Mean Force) in biomolecules. It works by breaking long pulling processes into a series of stages or windows along a specific reaction coordinate. adaptive steered molecular dynamics A minimal basis set in which each Slater-type atomic orbital is approximated by a fixed combination of 3 Gaussian functions. Gaussian functions are used because the integrals they need are much faster to compute than those of Slater-type orbitals. It is by far the most used member of the family and appears in almost every textbook example. (UNVERIFIED) STO-3G A minimal basis set in which each Slater-type atomic orbital is approximated by a fixed combination of 4 Gaussian functions. Gaussian functions are used because the integrals they need are much faster to compute than those of Slater-type orbitals. More Gaussian functions give a better approximation but no extra flexibility, because the number of orbitals is unchanged. (UNVERIFIED) STO-4G A process that turns molecular structures, trajectories and simulation data into pictures or animations. It is used to check a system before a run, to follow a run while it proceeds, and to understand the results afterwards. Some errors in a prepared system, such as a molecule placed in the wrong position, are easier to see in a picture than in numerical checks. (UNVERIFIED) visualization A widely used package of biomolecular simulation software, encompassing both a suite of programs (AmberTools) for simulation preparation, analysis, and modeling, and a high-performance simulation engine (pmemd) for running MD and free energy calculations. AMBER is associated with a popular family of force fields (e.g., ff14SB, GAFF) and is extensively used for simulating proteins, nucleic acids, lipids, and organic molecules. It includes capabilities for classical MD, QM/MM, and various advanced simulation techniques. AmberMD AMBER The freely available component of the AMBER software package, containing numerous programs and utilities for biomolecular simulation tasks excluding the main high-performance simulation engine (pmemd). AmberTools includes programs for system setup (tleap), trajectory analysis (cpptraj), structure preparation (antechamber), energy minimization and basic MD simulation (sander), QM calculations (sqm), free energy analysis (mmpbsa), and many others. It provides the essential infrastructure for preparing, analyzing, and modeling simulations run with AMBER or other programs. AmberTools A shell resource file, created when AmberTools or AMBER is installed, that sets up the shell environment for the programs, including the AMBERHOME or PMEMDHOME environment variable. It must be sourced rather than executed, and is often added to the login file so that the environment is ready in every new shell. A similar file, amber.csh, serves C-type shells. (UNVERIFIED) amber.sh A software program that decides which acidic and basic groups in a structure carry a proton at a stated acidity. It predicts this from the local environment of each group, since experimental structures rarely show hydrogens. The answer changes the charge of the system and so changes everything a simulation does. (UNVERIFIED) protonation tool A widely used software tool (available as a web server and standalone package) that automates many common tasks in preparing biomolecular structures (from PDB files) for continuum electrostatics calculations or simulations. Its core functions include adding missing heavy atoms, adding and optimizing hydrogen atom positions, assigning charges and radii from a chosen force field, and determining residue protonation states at a specified pH using the PROPKA algorithm internally. It outputs the structure in PQR format (PDB with Charge and Radius). PDB2PQR A protonation tool specifically focused on determining the most likely ionization and tautomeric states of small molecule ligands, often in the context of drug discovery or protein-ligand binding studies. Predicting ligand protonation can be challenging due to diverse functional groups and potential binding site environmental effects. ligand protonation tool A protonation tool designed for general use on biomolecular structures, typically proteins or nucleic acids obtained from sources like the PDB, to add hydrogens and determine appropriate protonation states for titratable residues at a given pH. These tools often focus on standard amino acids and sometimes common cofactors. PDB2PQR is a classic example of such a tool. generic protonation tool A software program in the Schrödinger suite that predicts which protonation and tautomeric forms of a small molecule are present at a given acidity, and how much each is favoured. It is run before docking or simulation, since the form chosen determines the molecule's charge and hydrogen bonding. It reports several forms with penalties rather than a single answer. (UNVERIFIED) Epik A computational tool specifically designed to generate realistic initial configurations of lipid bilayers, often including embedded proteins, for use in membrane simulations. These tools arrange lipid molecules (and potentially cholesterol, solvent) according to specified lipid types, compositions, and area per lipid, ensuring proper packing density and avoiding steric clashes, which is challenging to do manually. Examples include PACKMOL-Memgen, CHARMM-GUI's Membrane Builder, and InflateGRO. (UNVERIFIED) membrane packing tool A membrane packing tool in AmberTools that builds simulation-ready systems of a lipid bilayer, with or without an embedded protein, together with water and ions. It orients the protein in the membrane with Memembed and uses Packmol to pack the lipids, and handles complex lipid mixtures, different compositions in each leaflet and systems with several bilayers. It can also run simple AMBER steps on the result, such as parameterization and minimization. (UNVERIFIED) PACKMOL-Memgen A computational tool designed to predict or determine the likely orientation and transmembrane embedding depth of a protein structure within a lipid bilayer. Correctly positioning a membrane protein in the bilayer is a critical first step for setting up realistic membrane protein simulations. These tools often use algorithms based on hydrophobicity analysis, sequence features, or structural properties to find the most favorable position relative to the hydrophobic core and hydrophilic headgroup regions of the membrane. membrane protein orientation tool A specific computational tool or algorithm designed to predict the optimal orientation and insertion depth of a membrane protein within a lipid bilayer. It works from a knowledge-based statistical potential. memembed A simulation method classified by how much structural detail it keeps, from every atom represented separately to many atoms grouped into 1 particle. The choice is the main control on what a simulation costs and on what can be seen in its result. A grouped model runs far longer and further, but nothing smaller than a group can be observed in it. simulation type resolution-based simulation method All-atom molecular dynamics (AA-MD) is a simulation technique where every individual atom in the molecular system, including hydrogen atoms and explicit solvent molecules, is represented. The motion of each atom is simulated over time by numerically integrating Newton's equations of motion using forces derived from a classical force field. This approach provides high structural detail but is computationally intensive, limiting accessible timescales and system sizes. AA-MD all-atom molecular dynamics A model specifying which contributions make up the total potential energy of a system and what functional form each takes. It is the structure of a force field, as distinct from the numbers filling it. Two force fields with the same structure and different numbers behave differently; 2 with different structures cannot share numbers at all. (UNVERIFIED) interaction energy model A component of the potential energy stored in the stretching or compression of covalent bonds away from their preferred length. It is usually modelled as a spring, whose energy rises ever more steeply the further the bond departs from that length, so the bond can never break. That is why an ordinary force field cannot describe a chemical reaction. (UNVERIFIED) bond interaction energy A potential energy arising from the gravitational attraction between masses. Between 2 protons it is weaker than their electrostatic repulsion by about 36 orders of magnitude, so molecular simulations leave it out. It is 1 of the 4 fundamental interactions of physics, together with the electromagnetic, strong and weak interactions. (UNVERIFIED) gravitational interaction energy A potential energy arising from electromagnetic forces, covering both the attraction and repulsion between charges and the weaker interactions between magnetic moments. Every chemical interaction is ultimately of this kind, including bonding and dispersion. In a force field it appears explicitly as the electrostatic term and implicitly in all the others. (UNVERIFIED) electromagnetic interaction energy A potential energy arising from the strong and weak nuclear forces, which act only within an atomic nucleus over distances of about a femtometre. They hold the nucleus together and govern radioactive decay. They never appear in a molecular simulation, which treats each nucleus as a single unchanging particle. (UNVERIFIED) nuclear interaction energy A free energy analysis method that combines histograms from several biased simulations, such as the windows of an umbrella sampling run, into a single unbiased free energy profile. It solves self-consistently for the free energy offset of each window, which is what removes the bias. It needs the sampled ranges of neighbouring windows to overlap. (UNVERIFIED) WHAM weighted histogram analysis method analysis The Potential of Mean Force (PMF) method calculates the free energy profile of a system as a function of one or more specific coordinates, often called reaction coordinates or order parameters. It essentially describes the effective potential energy governing the system's behavior along those chosen coordinates, including entropic effects averaged over all other degrees of freedom. PMFs are crucial for understanding reaction mechanisms, binding processes, and conformational changes. potential of mean force analysis A free energy analysis method distinguished by how many simulations its snapshots come from: 1 simulation of the complex, or separate simulations of the complex, the free receptor and the free ligand. The single-trajectory choice cancels internal energy terms and is cheaper, but it cannot capture conformational change on binding. The choice affects both cost and accuracy, so it must be recorded with the result. (UNVERIFIED) trajectory approach analysis The single trajectory approach, often used in MM/PBSA and MM/GBSA calculations, involves running only one molecular dynamics simulation of the protein-ligand complex. Snapshots of the complex, the protein alone, and the ligand alone are then extracted directly from this single trajectory for endpoint free energy calculations. This reduces the computational cost compared to running three separate simulations but assumes the conformations sampled in the complex are representative for the free states. single trajectory A computational tool specifically designed to predict the pKa values of ionizable groups (acidic or basic sites) within a molecule, typically amino acid residues in a protein or functional groups in a ligand. The pKa value indicates the pH at which a group is 50% protonated/deprotonated. These tools use various methods, ranging from simple empirical rules to sophisticated calculations involving continuum electrostatics (solving Poisson-Boltzmann equation) or QM methods, to estimate how the molecular environment shifts the intrinsic pKa of a group. PROPKA is a well-known example. pKa tool A pKa tool that predicts the pKa values of ionizable groups in proteins and protein-ligand complexes from their 3-dimensional structure, using fast empirical rules. It estimates how far each group's pKa shifts from its value in water because of nearby hydrogen bonds, burial away from the solvent and interactions with charged groups. The results are used to choose the protonation state of each residue at the pH of a simulation, and it is built into other tools such as PDB2PQR. (UNVERIFIED) PROPKA A molecular mechanics Poisson-Boltzmann surface area tool, distributed as a Python program with AmberTools, that estimates binding free energies from snapshots of a simulation trajectory by the MM/PBSA or MM/GBSA method. For each snapshot it computes the molecular mechanics energy, the electrostatic solvation energy from a Poisson-Boltzmann or generalized Born model and a nonpolar term from the solvent accessible surface area, then averages them. It automates a popular but approximate way of estimating binding free energies. (UNVERIFIED) MMPBSA.py A computational tool that allows researchers to simulate the effects of mutations (changing one amino acid residue to another) on the structure, stability, dynamics, or binding affinity of a protein. These tools typically take a protein structure, allow the user to specify mutations, and then use algorithms (often based on molecular mechanics energy functions, rotamer libraries, or machine learning) to model the structural changes and estimate the impact on stability (ddG) or other properties. in-silico mutagenesis tool An in-silico mutagenesis tool that estimates how much a mutation changes the folding stability of a protein, using an empirical energy function applied to the protein's 3-dimensional structure. The energy function adds up terms for van der Waals contacts, solvation, hydrogen bonds, electrostatics and entropy, weighted to match measured stability data. It is widely used to predict the effect of single or multiple point mutations and in protein design. (UNVERIFIED) FoldX A software program that finds and measures the channels, pores and tunnels that run through a macromolecular structure, such as the pore of an ion channel or a tunnel leading to a buried enzyme active site. These pathways control how ions, water, substrates and products move into and out of a protein. Programs such as HOLE and CAVER find them from atomic coordinates. (UNVERIFIED) tunnel predictor tool A software tool for identifying and analyzing tunnels and channels in protein structures. CAVER uses computational geometry algorithms to explore the void space within a protein (or dynamic snapshots from a simulation) and simulates the introduction of probes to map the surface and the cavities of a target. It provides information on the location, dimensions (bottleneck radius), curvature, and physicochemical properties of identified tunnels. CAVER An open-source software suite for classical molecular dynamics simulation, designed mainly for biomolecules such as proteins, lipids and nucleic acids, and known for its high performance on both central and graphics processors. It supports many force field families, including AMBER, CHARMM, GROMOS and OPLS, a range of enhanced sampling and free energy methods, and many tools for preparing systems and analysing trajectories. It is widely used in academic and industrial research. (UNVERIFIED) GROMACS A versatile and long-established software package for molecular simulation, encompassing molecular mechanics (MM) and dynamics (MD), as well as capabilities for QM/MM and complex modeling tasks (stands for Chemistry at HARvard Macromolecular Mechanics). CHARMM is associated with its own widely used force field family (e.g., CHARMM36) and provides extensive functionalities for preparing, running, and analyzing simulations, particularly for proteins, lipids, nucleic acids, and carbohydrates. Both a free academic version and a licensed version exist. CHARMM The set of numbers defining how an ion interacts non-covalently in a force field, being its charge, its size and its interaction strength. Ions are hard to parameterise because a single set cannot match every measured property at once. The choice therefore depends on what the simulation is meant to reproduce. (UNVERIFIED) ion parameterization A set of Lennard-Jones parameters for single-atom ions, from monovalent to tetravalent, developed by Li, Merz and co-workers for the standard 12-6 form of the potential and fitted separately for several water models. Separate versions reproduce the hydration free energy, the ion-oxygen distance, or a compromise between them, because the 12-6 form cannot reproduce these properties at once for highly charged ions. AMBER also offers a normal-usage set that takes, for each ion charge, the version that works best. (UNVERIFIED) Li and Merz 12-6 An analysis tool specifically focused on reading, processing, manipulating, and analyzing molecular dynamics trajectory files. These tools provide functionalities to read various trajectory formats, select specific atoms or frames, perform calculations over time (like RMSD, distances, angles), calculate average structures, remove solvent, re-image periodic boundaries, and write out modified trajectories or extracted data. trajectory processing tool A powerful command-line trajectory analysis program, distributed with AmberTools, capable of processing trajectories from many simulation packages (AMBER, GROMACS, NAMD, CHARMM, etc.) and performing a wide variety of analyses. Its capabilities include calculating structural metrics (RMSD, radius of gyration, distances, angles, dihedrals), analyzing hydrogen bonds, performing principal component analysis (PCA), clustering, calculating interaction energies, grid analysis, and extensive data manipulation and output options. It is a central tool for analyzing simulation data in the AMBER ecosystem and beyond. CPPTRAJ An algorithm used in data analysis to group a set of objects (e.g., molecular conformations, data points) such that objects in the same group (cluster) are more similar to each other than to those in other groups. Clustering algorithms use various criteria and methods (e.g., distance-based, density-based, hierarchical) to identify inherent structures or patterns within unlabeled data. They are essential tools for exploring and understanding complex datasets. (UNVERIFIED) clustering algorithm A clustering algorithm that builds a hierarchy of clusters by starting with each data point as its own cluster and repeatedly merging the 2 closest clusters until a single cluster containing all points remains. This process creates a tree-like structure (dendrogram) that can be cut at different levels to obtain different numbers of clusters. It provides a nested partitioning of the data based on similarity. (UNVERIFIED) hierarchical agglomerative bottom-up algorithm A modelling method that predicts how molecules will bind from their fixed structures alone, without simulating them moving. It scores how well 2 shapes fit together and how favourable the contacts are, rather than following the binding as it happens. It is fast enough to test millions of candidate molecules, which no simulation-based method can do. (UNVERIFIED) static molecular interaction prediction modeling A computational method that predicts the structure of a complex formed when 2 proteins bind, given the structure of each on its own. It searches over the ways the 2 surfaces can meet and scores each arrangement. It is harder than docking a small molecule, because the contact area is far larger and both partners may change shape. (UNVERIFIED) protein-protein docking analysis A software program that estimates a binding or solvation free energy from the initial and final states of a process alone, without simulating any path between them. It combines molecular mechanics energies with a continuum solvation model and a surface area term. It is much cheaper than an alchemical calculation and correspondingly less rigorous. (UNVERIFIED) endpoint free energy tool A specific program, typically distributed as part of the AmberTools suite, designed to automatically assign atom types and calculate partial atomic charges for organic molecules, preparing them for use with molecular mechanics force fields like GAFF (General AMBER Force Field). Antechamber reads common molecular file formats and generates topology and parameter files needed for subsequent energy minimization or molecular dynamics simulations using programs like SANDER or PMEMD from the AMBER package. It is a crucial tool for preparing non-standard residues or ligands for simulation. Antechamber A system setup tool that writes the molecular topology file and the initial coordinate file for a system to be simulated. The topology lists the atom types, bonds, angles, dihedrals, charges and non-bonded parameters that the force field assigns, and the program builds it from a structure file and force field definition files. An example is tleap from AmberTools. (UNVERIFIED) topology and coordinate generation tool A topology and coordinate generation tool in AmberTools, the command-line version of the LEaP program, that builds the topology and initial coordinate files for AMBER simulations. It reads force field files, residue template libraries and user structure files such as PDB files, joins units, adds solvent and counterions, and writes the prmtop topology and inpcrd coordinate files that sander and pmemd read. It is the standard system-building step in the AMBER workflow. (UNVERIFIED) LEaP tleap Binding free energy encompasses computational techniques specifically designed to estimate the Gibbs free energy change (ΔGbind) associated with the non-covalent binding of two molecules. Calculating binding free energy is crucial for predicting molecular interactions, understanding binding affinity in drug design, and validating docking poses. These methods range from approximate endpoint methods to more rigorous alchemical transformations. binding free energy calculation analysis A free energy analysis method that estimates a binding free energy from snapshots of the bound and free states, computing the polar solvation term with the Poisson-Boltzmann equation. Each free energy is the sum of 3 terms: the molecular mechanics energy in vacuum, the polar solvation free energy and a non-polar solvation term scaled from the solvent-accessible surface area. It uses only the end states, so it is much cheaper than an alchemical calculation and correspondingly less rigorous. (UNVERIFIED) MM/PBSA MMPBSA molecular mechanics Poisson Boltzmann surface area analysis A free energy analysis method that estimates a binding free energy in the same way as the Poisson-Boltzmann variant, but computes the polar solvation term with the generalized Born approximation instead of solving the Poisson-Boltzmann equation. The approximation is much faster and usually close enough for ranking related ligands. The molecular mechanics and surface-area terms are unchanged. (UNVERIFIED) MM/GBSA MMGBSA molecular mechanics generalized Born surface area analysis A free energy analysis method that estimates a binding free energy from the change in the average electrostatic and van der Waals interaction energies of a ligand with its surroundings between the bound and the free state. The 2 changes are combined with empirical scaling coefficients rather than computed from first principles. It needs 1 simulation of the complex and 1 of the ligand in solution, which is far fewer than an alchemical route. (UNVERIFIED) LIE linear interaction energy analysis A free energy analysis method that computes a free energy difference by simulating a reversible path between 2 states through intermediates that need not be physically realisable. The energy function is changed gradually along the path, for example turning a ligand's interactions off, and the free energy is recovered from how the system responds. Free energy perturbation and thermodynamic integration are the 2 standard ways of doing it. (UNVERIFIED) alchemical free energy analysis Free Energy Perturbation (FEP) is an alchemical method for calculating free energy differences between two states (A and B) by simulating a series of intermediate states that gradually transform A into B. The free energy change between adjacent intermediate states (λ and λ+Δλ) is calculated using the Zwanzig equation, which involves ensemble averaging the energy difference exponentiated. Summing the free energy changes over all intermediate steps yields the total free energy difference between A and B. FEP free energy perturbation analysis The Channel Annotation Package (CHAP) is a specialized software tool used to analyze the functional properties of pores and channels in membrane proteins. It calculates a detailed pore radius profile combined with a hydrophobicity profile to estimate the free energy barriers for water and ion permeation. This analysis allows researchers to visualize solvent behavior inside the pore and determine the conductive state of ion channels. CHAP Binding Free Energy Estimator 2 (BFEE2) is a comprehensive computational workflow designed to calculate protein-ligand binding free energies with 'chemical' accuracy. Its packages automate a rigorous multi-step process that includes enhanced sampling of the ligand's binding pose and the use of alchemical or geometrical free energy methods to compute the final affinity. BFEE2 binding free energy estimator 2 analysis A free energy analysis method that recovers an equilibrium free energy difference from simulations deliberately driven away from equilibrium, by applying a fluctuation theorem to the work done. Many fast, irreversible transformations are run between the 2 states and the work distribution is analysed with Jarzynski's equality or Crooks' theorem. It can be cheaper than the equilibrium alchemical route because each transformation is short. (UNVERIFIED) NEQ non-equilibrium free energy analysis An implicit solvent model that estimates the solvation free energy as the sum of an electrostatic term, from a numerical solution of the Poisson-Boltzmann equation, and a non-polar term proportional to the solvent-accessible surface area. The surface area term stands for the cost of making a cavity in the solvent and for the van der Waals attraction between solute and solvent. It is often applied to snapshots from a simulation to estimate binding free energies. (UNVERIFIED) PBSA Poisson-Boltzmann surface area An implicit solvent model that estimates the solvation free energy as the sum of a generalized Born electrostatic term and a non-polar term proportional to the solvent-accessible surface area. It works like the Poisson-Boltzmann surface area approach, but replaces the Poisson-Boltzmann calculation with the much cheaper generalized Born approximation. This makes it fast enough to use during simulations or for large numbers of structures, though its accuracy depends on the generalized Born variant used. (UNVERIFIED) GBSA generalized Born surface area A defined collection or list that assigns specific atomic radius values to different atom types (defined by element, hybridization state, and sometimes chemical environment). These radii are used in various computational molecular modeling contexts, such as calculating molecular surfaces (like Van der Waals surface or Solvent Accessible Surface Area - SASA), estimating molecular volumes, defining cavities, or as parameters within implicit solvent models (e.g., determining effective Born radii for GB models or defining the dielectric boundary for PB models). Different radii sets exist (e.g., Bondi Van der Waals radii, radii optimized for solvation calculations like PARSE), derived from different physical principles or empirical fitting procedures. (UNVERIFIED) radii set A set of van der Waals radii for the common elements, published by Bondi in 1964 and derived from crystallographic contact distances and gas-phase properties. Many later radii sets, such as the modified Bondi sets used with generalized Born models, were derived from it. It was not fitted for use in any solvation model. (UNVERIFIED) bondi A set of van der Waals radii derived from the Bondi set with adjustments for particular atom types, chiefly hydrogen. It is the default radii set in the AMBER LEaP program and the one recommended for the Hawkins-Cramer-Truhlar generalized Born model (igb=1). The prefix stands for modified. (UNVERIFIED) mbondi A set of van der Waals radii that is the second modification of the Bondi set in AMBER, introduced with the generalized Born model of Onufriev, Bashford and Case. It is the radii set recommended for that model (igb=2 and igb=5), and for nucleic acids it is also used with the GB-Neck2 model. Like the other modified Bondi sets, it changes the radii of some atoms, chiefly hydrogens, to improve generalized Born solvation energies. (UNVERIFIED) mbondi2 A thermodynamics analysis method that estimates the entropy term of the free energy from the vibrational modes of a system. It approximates entropy from how the system vibrates about an average structure, rather than counting states directly. Normal mode analysis and the quasiharmonic approximation are the 2 usual routes. (UNVERIFIED) enthropic contributions mode analysis A dynamics analysis method that estimates vibrational entropy from a trajectory by treating atoms as oscillating harmonically about their average positions. It diagonalises the mass-weighted covariance matrix of the atomic fluctuations to obtain effective frequencies, which are then put through the standard harmonic oscillator formulas. Unlike normal mode analysis it uses the sampled motion rather than a single minimum, so it captures anharmonic breadth as an effective frequency. (UNVERIFIED) QHA quasiharmonic analysis A dynamics analysis method that computes the vibrational modes and frequencies of a molecule by treating its motion as harmonic about a single energy minimum. It builds the matrix of second derivatives of the potential energy at that minimum and diagonalises it, giving collective modes as eigenvectors and their frequencies from the eigenvalues. Those frequencies are what a vibrational entropy estimate is built from. (UNVERIFIED) NMA normal mode analysis A trajectory approach for end-point binding free energy calculations in which only the complex is simulated, and the receptor and ligand structures are taken from that same trajectory. Because every structure comes from 1 simulation, the internal energies of the parts cancel exactly, which removes most of the statistical noise. The price is that neither part can relax into the shape it would take on its own, so any real conformational change on binding is missed. (UNVERIFIED) 1 trajectory approach A trajectory approach for end-point binding free energy calculations in which the complex, the free receptor and the free ligand are each simulated separately. Each component is therefore sampled in the state it is actually in, which captures any conformational change on binding. The cost is 3 simulations instead of 1, and much larger statistical noise, because the internal energies no longer cancel. (UNVERIFIED) 3 trajectory approach A type of utility program used to help set up a thermodynamic integration (TI) free energy calculation. It assesses the similarity between the starting and ending molecules of a planned alchemical transformation, often by calculating the overlap of their potential energies. The results from this tool are used to design an optimal pathway with an appropriate number of intermediate steps, which is crucial for ensuring the TI calculation will be accurate and reliable. (UNVERIFIED) thermodynamic integration similarity calculation tool A software program that compares molecules by overlaying their 3-dimensional shapes and chemical features rather than their bond connectivity. Two molecules built on different scaffolds can occupy the same space and bind the same site, which a structure comparison would miss. It is used to search for such replacements. (UNVERIFIED) rapid overlay of chemical structures ROCS A workflow program for AMBER and AmberTools that automates the calculation of binding free energies for a set of ligands that bind to the same receptor. It sets up and analyses the simulations for 3 approaches: MM-PBSA and MM-GBSA, linear interaction energy, and thermodynamic integration. It is written in Perl and calls AmberTools programs such as tleap, antechamber and cpptraj. (UNVERIFIED) AMBER-FEW An open-source software library for thermodynamic calculations using the CALPHAD method, which models phase equilibria in multi-component systems from fitted Gibbs energy expressions. It is used in materials science and metallurgy to compute phase diagrams. It is a thermodynamic modelling tool rather than a molecular simulation one. (UNVERIFIED) OpenCalphad A Monte Carlo simulation program, open source and developed at Delft University of Technology, that computes phase and reaction equilibria with the continuous fractional component method. In that method a fractional molecule, whose interactions are scaled, is inserted or removed gradually, which makes molecule exchanges efficient even in dense liquids. Chemical potentials and partial molar properties can then be obtained directly from a single simulation. (UNVERIFIED) Brick-CFCMC A free energy analysis method defined by how a thermodynamic integration calculation is arranged: how the path between the 2 states is built, how intermediate states are spaced and sampled, and how the atoms of the 2 states are matched. These choices affect cost and convergence but not the quantity being computed. Examples are the single-topology, dual-topology and multi-step approaches. (UNVERIFIED) thermodynamic integration approach A thermodynamic integration approach in which independent simulations are run at a set of fixed values of the coupling parameter and the results integrated numerically. Each simulation is independent of the others, so they can all be run at once. It is the arrangement every other variant is compared against. (UNVERIFIED) standard thermodynamic integration A thermodynamic integration approach in which the average derivative of the energy with respect to the coupling parameter is predicted at coupling values that were not simulated, by reweighting data stored during the simulations that were run. This fills in the integrand between simulated points without extra simulations. It can also be used to compare other ways of making the energy depend on the coupling parameter and to choose one that gives a smoother curve. (UNVERIFIED) extended thermodynamic integration A thermodynamic integration approach in which the atoms of both states are present in the system at once, with each set coupled to the rest according to the coupling parameter and invisible to the other. It handles changes in connectivity that a single set of atoms cannot represent. It costs more, because both states are simulated throughout. (UNVERIFIED) dual-topology approach A thermodynamic integration approach in which 1 set of atoms has its types, charges and van der Waals parameters changed along the path, with atoms present in only 1 state turned into, or grown from, non-interacting dummy atoms. It keeps the atom count fixed throughout. It is compact but awkward when the 2 states have different connectivity. (UNVERIFIED) single-topology approach A thermodynamic integration approach that uses a soft-core potential to cover the whole transformation in 1 window, or very few. The soft-core form keeps the energy finite as atoms appear and disappear, which is what makes so few windows possible. It saves simulation time at the cost of relying heavily on the soft-core form being well behaved. (UNVERIFIED) 1-step soft-core approach A thermodynamic integration approach that divides the path between the 2 states into several discrete windows, each sampled separately. Dividing it keeps the change between neighbouring windows small enough for the samples to overlap. It is the standard arrangement, and the alternative of a single window is a special case. (UNVERIFIED) multi-step approach A thermodynamic integration approach in which the simulations at different coupling values run in parallel and periodically swap configurations, with the swap accepted or rejected by a Metropolis criterion. Swapping lets a configuration travel along the whole path, which relieves sampling bottlenecks at any single value. It often converges faster than independent windows for the same total simulation time. (UNVERIFIED) Hamiltonian replica exchange A thermodynamic integration approach in which an even number of coupled replicas, spaced equally along a closed path from 1 end state to the other and back, move along the coupling parameter together as a single variable advances. At any moment half of the replicas move towards 1 end state and half towards the other, like items on a conveyor belt. The barriers felt by the replica set shrink as replicas are added, so the whole range of the coupling parameter is covered almost evenly. (UNVERIFIED) CBTI conveyor belt thermodynamic integration A thermodynamic integration approach that chooses the number and the spacing of the intermediate coupling values so as to minimise the statistical error for a given amount of simulation. Even spacing is wasteful, because the integrand changes much faster near the end points than in the middle. The spacing may be fixed in advance or adjusted as the calculation runs. (UNVERIFIED) optimization of alchemical pathways A workflow that sets up, runs and analyses a free energy calculation as one automated sequence. It manages the many coupling states, their simulations and the collection of their output, which is otherwise done by hand and easily got wrong. It is what makes a campaign over many ligands practical. (UNVERIFIED) FEW free energy workflow A thermodynamic integration analysis that estimates the free energy integral from simulations at or near the end states, or at very few intermediate values of the coupling parameter, by extrapolating the integrand between them. This avoids simulating many intermediate states and so greatly reduces cost. The result is only as accurate as the extrapolation is for the system studied. (UNVERIFIED) thermodynamic integration extrapolation approach A thermodynamic integration approach that samples the integrand only at or near the 2 end points and assumes it varies linearly between them. Assuming linearity removes every intermediate simulation, which is where nearly all the cost lies. It is only valid where the integrand really is close to linear, which is uncommon. (UNVERIFIED) thermodynamic integration with linear extrapolation Analysis techniques and calculated properties that are particularly characteristic of or frequently applied to data generated from molecular dynamics (MD) simulations, especially those aimed at calculating free energies or characterizing structural dynamics. These methods help interpret the complex time-series data produced by MD. (UNVERIFIED) free energy and molecular dynamics analysis A free energy analysis method that estimates the free energy difference between 2 thermodynamic states by combining the forward and reverse work distributions sampled in both. For a given amount of simulation data it gives the lowest-variance estimate available. It is the 2-state case of the multistate Bennett acceptance ratio. (UNVERIFIED) BAR Bennett acceptance ratio analysis A free energy analysis method that combines data from simulations run at many thermodynamic states at once, and returns the free energy difference between any pair of them together with equilibrium averages. It uses every sample from every state rather than only neighbouring pairs, which is what makes it statistically optimal. It generalises the 2-state Bennett acceptance ratio. (UNVERIFIED) MBAR multi-state Bennett acceptance ratio analysis A mathematical analysis that determines whether a set of vectors or functions is linearly dependent, meaning that 1 of them can be written as a combination of the others. In quantum chemistry it is used to find basis functions that are almost combinations of the others, which make a calculation numerically unstable and are usually removed. In trajectory analysis it can show that a set of collective variables contains redundant ones. (UNVERIFIED) linear dependence evaluation analysis An experimental method is a procedure carried out in a laboratory to make observations and gather data about the real world. As a type of method, it provides real-world data, such as a structure from X-ray crystallography or dynamics from nuclear magnetic resonance, which serves as a crucial input for validation of a computational simulation. experimental method A modeling method is a computational technique used to to predict, refine, and simulate 3D molecular arrangements. As a type of computational method, it typically involves predicting a static structure or interaction pose without simulating its dynamic behavior over time, such as in protein-ligand docking. These methods often employ search algorithms and scoring functions to evaluate and rank potential conformations based on criteria like shape complementarity and interaction energy. modeling method A software program built for statistical analysis, data handling, plotting and reporting. It is not specific to molecular simulation, but simulation output is data like any other and is often analysed in one. R and similar environments are the usual examples. (UNVERIFIED) statistical package A free software environment and programming language for statistical computing and graphics. Users worldwide contribute packages to it, covering most areas of statistics, data visualization and data science. In simulation analysis it is used to plot results such as energy or root mean square deviation time series, to run statistical tests on sampled properties and to cluster conformations. (UNVERIFIED) R A computational method that generates a series of states of a molecular system by repeatedly applying physical rules to its current state, either advancing it through time as in molecular dynamics or making random moves as in Monte Carlo sampling. It produces a trajectory, or a set of sampled states, rather than a single answer. The choice of method fixes which physical laws apply and how each new state is produced. (UNVERIFIED) simulation method A thermodynamics analysis method that estimates the contribution of disorder to the free energy of a molecular system. It works from the system's vibrational modes or from the spread of conformations in a trajectory, most often through normal mode analysis or the quasiharmonic approximation. Multiplied by the temperature, the result is the term that is subtracted from an energy to give a free energy. (UNVERIFIED) entropy analysis OBSOLETE. A plan specification is a document or set of rules that describes how a study or experiment should be designed and carried out. As a type of information content entity, it represents the formal design of a computational experiment, such as the ligand pairs selection criteria for a relative binding free energy study. This specification outlines the rationale and criteria for selecting systems and parameters before the simulation or analysis is performed, ensuring a systematic and reproducible approach. This concept is often described in publications in sections titled "Experimental Design," "Study Protocol," "Selection Criteria," or "Computational Strategy." Kept as a MOLSIM term rather than reusing IAO:0000104 plan specification, which is already present in the IAO import module but referenced nowhere. Reusing it would pull in two grouping levels MOLSIM has no use for, IAO:0000033 directive information entity and IAO:0000030 information content entity, and would place this branch under the BFO spine (BFO:0000031 generically dependent continuant and above), which MOLSIM defers. Revisit only if upper-ontology alignment is opened. Retired 2026-08-12 as part of the COB alignment. IAO:0000104 plan specification says the same thing and is reached from IAO:0000030 information content entity, which MOLSIM already adopted. This retirement waited until the branch was cleaned: MOLSIM:000310 previously held four children of which three were not plans, and those moved to MOLSIM:002251 atom selection specification and MOLSIM:002252 selection criteria. Its one remaining child, MOLSIM:002155 simulation protocol, was re-parented onto IAO:0000104 in the same change. obsolete plan specification true A calculation that measures how much a molecule's shape has changed compared to a reference structure by quantifying the root-mean-square deviation of atomic coordinates after optimal superposition. This can be performed against a single static starting structure over time (1D trajectory), or calculated for every pair of frames in a simulation to produce a frame-to-frame comparative matrix, often displayed as a 2D heat map (2D RMSD analysis). It is the most common procedure for assessing simulation stability, monitoring conformational transitions, and serving as a metric for clustering algorithms. RMSD analysis RMSF analysis is a calculation that identifies which parts of a molecule are the most flexible or wobbly. As a type of structural analysis, this method quantifies the mobility of individual atoms or residues by calculating the root-mean-square fluctuation of their positions around the time-averaged structure from a trajectory. The resulting per-residue RMSF values are directly comparable to crystallographic B-factors and are used to identify flexible loops, termini, or side chains. RMSF analysis Distance RMSD analysis is a way to compare two molecular shapes based on their internal distances, without needing to align them first. As a type of structural analysis, this method calculates the root-mean-square deviation of a matrix of internal, pairwise atomic distances between two conformations, providing a metric of structural similarity that is independent of overall rotation and translation. This is particularly useful for comparing highly flexible molecules or for clustering based on internal conformational changes rather than global alignment. distance RMSD analysis Radius of gyration analysis is a calculation that measures how compact or spread out a molecule is. As a type of structural analysis, this method quantifies the overall size and shape of a molecule or atomic selection by calculating the root-mean-square distance of its atoms from their common center of mass. Monitoring the radius of gyration over time is a standard procedure to track large-scale conformational changes, such as protein folding, unfolding, or domain movements. radius of gyration analysis A structural property consisting of the pair of backbone rotation angles for each residue in a protein. Plotted against each other across many residues, the pairs cluster into a few regions and leave most of the plot empty, because the remaining combinations put atoms on top of one another. A residue falling in an empty region is a sign that the structure is wrong there. (UNVERIFIED) Rama plot ramachandran plot data A potential of mean force is an energy profile that shows how much energy it takes to move a molecule from one place to another. As a type of free energy, the Potential of Mean Force (PMF) is an output_data_descriptor that represents the free energy of a system as a function of one or more specific geometric coordinates, known as reaction coordinates. The PMF describes the energetic landscape along a pathway, with its minima representing stable states and its maxima corresponding to transition state energy barriers. This is commonly referred to in analysis outputs and publications as "PMF," "free energy profile," "Potential of Mean Force," or "pmf.dat." PMF potential of mean force JSON (JavaScript Object Notation) is a textual data interchange format with a simple syntax and low parsing cost. It is human-readable and it has a flexible structure. The JSON format is widely used and it is natively supported for parsing and serialization in most mainstream programming languages. JSON format A text data format that records structured data mainly through indentation to show nesting, without the need for brackets or closing tags. It is easy for a person to read and edit, which is why configuration and workflow files use it. Its reliance on indentation also means a misplaced space changes the meaning. (UNVERIFIED) YAML format A standardized, minimalistic text-based file format used to store the Cartesian coordinates of atoms. It typically consists of a header line (atom count), a comment line, and one line per atom containing the element symbol and X, Y, Z positions. It may contain multiple frames (i.e. trajectory). The XYZ file format is widely supported by many programs, although many slightly different XYZ file formats coexist (Tinker XYZ, UniChem XYZ, etc.). standard xyz format A text data format for configuration files, built from key and value pairs grouped into named sections. Its rules are deliberately narrow, so a file has only one possible reading. That is its advantage over more flexible formats, where the same text can be interpreted more than one way. (UNVERIFIED) TOML format A binary data format that stores Python objects exactly as they exist in memory, so they can be written to a file and restored later. It preserves structures no general format can express. It can only be read back by Python, and reading one from an untrusted source can execute code, so it is unsuitable for exchange. (UNVERIFIED) pickle format A text-based data file format generated by GROMACS analysis tools, designed for plotting 1D or 2D graphs. It contains tabular numerical data alongside metadata and formatting commands intended to be read directly by the Grace (xmgrace) plotting program. xvg format An atom type that classifies an atom based strictly on its atomic number (number of protons in its nucleus), defining its fundamental chemical element. In structure files, it is typically represented by its standard IUPAC one- or two-letter chemical symbol (e.g., C, O, Fe). atomic element element type An atom type assigned by a specific molecular mechanics force field that dictates the atom's non-bonded and bonded interaction parameters. It encodes not only the element, but also the atom's hybridization state, local chemical environment, and connectivity (e.g., distinguishing between an sp3 aliphatic carbon and an sp2 aromatic carbon). (UNVERIFIED) mechanics atom type force field atom type A stochastic simulation method that relies on repeated random sampling to explore the configurational space of a molecular system. Unlike molecular dynamics, which integrates equations of motion over time, Monte Carlo generates a sequence of states by proposing random trial moves (such as atom displacements or molecular rotations) and accepting or rejecting them based on a thermodynamic probability criterion, typically the Metropolis algorithm. It is widely used for calculating equilibrium thermodynamic properties and sampling conformational ensembles. (UNVERIFIED) MC MC simulation monte carlo simulation A free energy analysis that estimates the free energy difference between 2 states from simulations of those 2 end states only, without simulating the intermediate states between them. It is cheaper but less rigorous than methods that sample a path between the states. MM/PBSA and MM/GBSA are common examples; they combine molecular mechanics energies with continuum solvent models, using snapshots of the bound and unbound states. (UNVERIFIED) endpoint free energy analysis Active site focusing is a strategy used in some endpoint free energy methods to improve the calculation by concentrating on the most important region of the system. This might involve performing the energy analysis only for residues within a certain distance of the binding site, or using a more detailed model for that region. For experts, this approach aims to reduce computational noise and focus the calculation on the energetic contributions that are most relevant to the binding event. active site focusing analysis A modelling method that prepares a molecular system for simulation by carrying out the steps needed to build a complete and chemically correct model, such as adding solvent molecules, placing ions and fixing the order of atoms. A careful setup gives a valid starting point that matches the conditions the simulation is meant to represent. (UNVERIFIED) system setup modeling An ion placement method is a specific procedure used during system setup to add ions, such as sodium or chloride, to the simulation box. The primary purpose is usually to neutralize the total charge of the system and to mimic a specific salt concentration. ion placement analysis An ion placement analysis that works out how many positive and negative ions to add to a simulation box to reach a target salt concentration. The calculation takes into account the volume of the box and the counter-ions already added to make the solute neutral. Getting it right makes the simulated system match the intended ionic strength. (UNVERIFIED) ion concentration calculation analysis A part of a force field, such as 1 of the energy terms it is built from, a parameter type, or an atom typing or residue definition that the energy terms use. The energy terms include bond stretching, angle bending, rotation about bonds, van der Waals and electrostatic terms, and their sum is the total energy. The parts of a force field are fitted together, so parts from different force fields usually cannot be mixed freely. (UNVERIFIED) force field component An optimization algorithm that helps the iterative self-consistent field calculation of a molecule's electronic structure reach a solution in which the electron density and the field it creates agree. Simple repetition of the self-consistent field step can oscillate between 2 near-solutions instead of settling, particularly for metals and open-shell systems. Methods such as extrapolation from previous iterations and level shifting damp or remove this oscillation. (UNVERIFIED) SCF convergence algorithm An SCF convergence algorithm that raises the energies of the unoccupied orbitals by a fixed amount during the iterations of a self-consistent field calculation. Widening the gap between occupied and unoccupied orbitals damps the mixing between them from one iteration to the next, which is what causes oscillation. The converged solution is the same as without the shift, so the shift changes only how quickly and reliably the calculation converges. (UNVERIFIED) level shifting algorithm A Tversky index algorithm that compares the chemical features of 2 aligned molecules, such as hydrogen bond donors and acceptors, with weights that emphasize the fit (database) molecule: α 0.05 and β 0.95. A high score means that the features of the fit molecule are largely contained in those of the reference (query) molecule. It complements shape-based similarity measures in molecular comparison workflows. (UNVERIFIED) fit color Tversky Restrained subset dynamics is a simulation method where restraining potentials are applied to only a specific part, or subset, of the molecular system. This allows the selected atoms to be held fixed or guided, while the rest of the system is free to move and respond dynamically. restrained subset dynamics A simulation method that combines 2 or more levels of description within one system, such as quantum mechanics for a reacting group and a force field for everything else. The expensive description is applied only where it is needed. The difficulty is always the boundary, where the 2 descriptions must be joined without introducing artefacts. (UNVERIFIED) multiscale simulation method A charge analysis that assigns partial charges to atoms by giving the electron population of each basis function to the atom it is centred on, and splitting the population shared between functions on 2 atoms equally between them. It requires no fitting and is very cheap to compute. Its values depend strongly on the basis set used, so they are unreliable for anything beyond a rough comparison. (UNVERIFIED) Mulliken charge analysis A pressure control algorithm that holds pressure steady by scaling all 3 box dimensions by the same factor, so the box changes size but not shape. It suits a system with no preferred direction, such as a protein in water. It is the wrong choice for a membrane, where the plane of the bilayer and the direction across it behave differently. (UNVERIFIED) isotropic position scaling barostat A numerical method that advances a differential equation forward in small steps, because no exact solution can be written down. The equations of motion for more than 2 interacting particles are of that kind, so every molecular dynamics simulation is doing this. The size of the step controls both the accuracy and the cost. (UNVERIFIED) numerical integration A class of mathematical techniques specifically used in molecular dynamics simulations to numerically solve the equations of motion governing particle movement. These algorithms approximate the positions and velocities of atoms at future discrete time steps based on the forces acting upon them at the current step. They are fundamental for generating the trajectory of the molecular system over time. (UNVERIFIED) integration analysis Auto-correlation analysis examines the correlation of a time series (a sequence of data points ordered in time, e.g., from a simulation) with lagged versions of itself. It helps identify repeating patterns, seasonality, or the characteristic timescale (memory) over which values in the series are statistically dependent. The primary output is the Auto-Correlation Function (ACF). auto-correlation analysis A dynamics analysis method that measures the statistical relationship between fluctuating quantities recorded during a simulation. It covers both how a single property relates to itself after a delay and how 2 different properties vary together or in opposition. The results reveal which motions are linked and over what timescales. (UNVERIFIED) correlation analysis An anisotropic position scaling barostat is a pressure control algorithm that allows the lengths of the three simulation box vectors to fluctuate independently to maintain a constant pressure tensor. This allows the shape of the simulation box to change, which is necessary for simulating systems that are not uniform in all directions, such as lipid membranes or solid crystals. anisotropic position scaling barostat A statistical test that judges whether 2 groups have different averages, without assuming they scatter by the same amount. It adapts Student's t-test, which assumes equal scatter in both groups. It suits comparisons between independent simulations whose results may scatter by different amounts, such as runs of different lengths or under different conditions. (UNVERIFIED) Welch's T-test analysis Auto-correlation coefficients are the numerical values that make up the Auto-Correlation Function (ACF) at different time points (τ). Each coefficient measures the linear correlation between the time series and the same series shifted by τ time steps. auto-correlation coefficients analysis An auto-correlation analysis that measures the correlation between a time series and itself a given number of steps back, after removing the linear effect of all the steps in between. Plain auto-correlation mixes a direct relationship with one carried indirectly through the intermediate points, and this analysis keeps only the direct part. It is used, for example, to choose the order of a time series model. (UNVERIFIED) partial auto-correlation coefficients analysis A statistical method that fits a relationship between an outcome and one or more explanatory quantities. It gives both the shape of the relationship and how much of the variation it accounts for. It is how a simulated quantity is checked against a measured one across a set of systems rather than one at a time. (UNVERIFIED) regression analysis A statistical method that judges whether an observed difference is larger than would be expected from sampling alone. It computes how likely the observed result would be if there were no real difference, and that likelihood is compared against a chosen threshold. It answers whether a difference is real, not whether it is large enough to matter. (UNVERIFIED) hypothesis testing analysis A statistical method that repeatedly draws new samples from the data already collected and recomputes a result, or refits a model, on each one. It is used to estimate how uncertain a result is, as in bootstrapping, or to test a model on data it was not fitted to, as in cross-validation. It replaces an assumption about how the data are distributed with computation. (UNVERIFIED) resampling method An experimental technique that measures the X-rays scattered at small angles by a sample in solution, and from that pattern infers the overall size and shape of the molecules in it. Because it works in solution it reports on the ensemble actually present, not on 1 crystallised conformation. It gives low-resolution shape rather than atomic positions. (UNVERIFIED) SAXS small angle x-ray scattering OBSOLETE. Similarity and divergence measures are mathematical quantities used to compare two probability distributions, datasets, or sequences, quantifying how similar or different they are. Similarity measures indicate closeness (higher values mean more similar), while divergence measures quantify the difference or 'distance' between them (higher values mean more different). These are used in various fields including information theory, statistics, and bioinformatics (e.g., comparing conformational ensembles). Retired 2026-08-13. It held a single child, MOLSIM:000118 Jensen-Shannon divergence analysis, which now sits directly under MOLSIM:000400 statistical analysis beside correlation analysis and hypothesis testing analysis. Three faults made it worth removing rather than renaming: the label joined two ideas with and, which reviewers raised as a pattern in August 2026; it grouped one term, so it explained nothing; and its definition described similarity and divergence measures, meaning quantities, while the class was labelled an analysis and parented under statistical analysis. The same shape as MOLSIM:000918 global structure analysis, retired on 2026-08-08. obsolete similarity and divergence analysis true The binpos ("binary position") format is binary file format from Amber that stores atomic coordinate frames (i.e. a trajectory) from a molecular dynamics simulation. The Scripps Research Institute is named upstream as the originator of the format. Recorded here rather than in the definition, which was verified on 2026-09-18 without it. Source: EDAM 1.25, class format_3885 BinPos (http://edamontology.org/format_3885). .binpos Scripps Research Institute BinPos Scripps binpos format binpos format A method used to check whether the results or models generated by biomolecular simulations are accurate, physically realistic, and reliable. These techniques assess the quality of simulation outputs by comparing them against known experimental data or established theoretical principles. Their goal is to ensure the simulation provides a meaningful representation of the molecular system being studied. (UNVERIFIED) validation analysis OBSOLETE. A type of statistical property that represents the probability of obtaining results at least as extreme as the observed results, assuming that the null hypothesis is true. A small p-value (typically < 0.05) is used to argue that the observed result is statistically significant. It is a fundamental concept in hypothesis testing. This is a standard output of statistical tests. Obsoleted 2026-07-29: duplicated STATO:0000700. STATO is the canonical OBO home for statistics; reused in place of this term. obsolete p-value true A free energy analysis method that computes the free energy difference between 2 states by simulating a series of intermediate states along a coupling parameter and integrating the average derivative of the energy with respect to that parameter. The path between the states does not have to be physically realisable, only reversible. It is 1 of the 2 standard alchemical methods, alongside free energy perturbation. (UNVERIFIED) TI thermodynamic integration analysis A thermodynamic integration approach in which the integrand is sampled at intermediate values of the coupling parameter rather than only at the end points, so the integral is computed from the sampled curve. Sampling the middle costs a simulation at each intermediate state and removes the assumption that the curve is a straight line. It is the arrangement nearly all thermodynamic integration uses. (UNVERIFIED) thermodynamic integration without linear extrapolation An analysis method that turns molecular data into pictures, such as 3-dimensional images of molecules, animations of simulation trajectories or 2-dimensional plots of computed quantities. Seeing the results helps researchers understand complex structures and communicate their findings. (UNVERIFIED) visualization analysis A worm plot is a visualization method that displays a collection of superimposed molecular structures from a simulation or an NMR ensemble. Each structure is typically represented as a smooth tube or "worm" tracing the path of the protein backbone. This method provides a clear visual impression of the conformational variability and dynamic fluctuations within the molecular ensemble. (UNVERIFIED) worm plot visualization analysis A helical density map is a visualization method used to represent the three-dimensional distribution of atoms within a helical structure, averaged over time from a simulation. It creates a 3D grid where the value at each point represents the probability of finding an atom, which can then be displayed as a contoured surface. This method is particularly useful for analyzing the average structure and packing of molecules like DNA or alpha-helical proteins. (UNVERIFIED) helical density map visualization analysis A computational method for placing ions, like sodium or chloride, into a simulation box containing a charged molecule like a protein. It works by calculating the electrostatic potential around the molecule and then using a mathematical operator, the Laplacian, to identify the most electrostatically favorable positions for the ions to be placed. This approach is more physically realistic than simply placing ions randomly, as it positions them in regions where they are most likely to be found, such as near oppositely charged amino acid residues. Laplacian mapping ion placement analysis An enhanced sampling simulation technique that enhances traditional biased molecular dynamics to calculate binding affinities, mechanical unfolding pathways, and free energy profiles (e.g., Potential of Mean Force) in biomolecules. It applies an evolving, position-dependent energy bias to help the system escape local minima along a specific reaction coordinate. adaptively biased molecular dynamics Driven adaptively biased molecular dynamics (d-ABMD) is an enhanced sampling technique that combines two approaches to explore a specific molecular process. It uses an adaptive biasing potential, like in metadynamics, to overcome energy barriers, while simultaneously applying an external "driving" force to guide the system along a predefined reaction coordinate. This dual approach is particularly effective for efficiently calculating the free energy profile of a complex conformational change that is too slow to occur spontaneously. driven adaptively biased molecular dynamics Redox potential replica-exchange molecular dynamics is a specialized version of the REMD method where the different parallel simulations are run at different electrochemical redox potentials, rather than different temperatures. This allows the system to efficiently sample different oxidation states in a simulation that is coupled to an electron bath. redox potential REMD redox potential replica-exchange molecular dynamics A restrained simulation is a type of simulation where an additional, artificial energy potential is applied to guide the system's motion. These restraints can be used to force a molecule to adopt a certain conformation, to prevent it from moving too far, or to incorporate experimental data into the simulation. Restraints are a powerful tool for targeted sampling, structure refinement, and calculating potentials of mean force. restrained simulation method An electron microscopy map restraint method is a simulation technique where the simulation is guided by a three-dimensional density map obtained from an electron microscopy experiment. An additional energy term is added that penalizes conformations where the simulated atoms do not fit well into the experimental density. For structural biologists, this method is a powerful tool for refining low-resolution experimental structures into high-resolution, physically realistic atomic models. electron microscopy map restraint method A numerical method that solves a differential equation by replacing each derivative with the difference between values at nearby grid points. The continuous problem becomes a large set of simultaneous equations, 1 per grid point. It is widely used to solve the Poisson-Boltzmann equation for a molecular shape. (UNVERIFIED) finite-difference analysis An analysis method that computes approximate numerical answers to mathematical problems that cannot be solved exactly, such as integrating the equations of motion or evaluating the integrals of statistical mechanics. Almost every step of a simulation depends on it, because for a real molecule these problems have no exact solution that can be written as a formula. A central concern is how large the approximation error is and whether it grows as the calculation goes on. (UNVERIFIED) numerical analysis A numerical method for solving a differential equation across a boundary where the solution or its derivative jumps, such as the surface between a solute and continuum solvent. It modifies the finite-difference stencil near the boundary to take the known jump conditions into account, instead of smoothing over them. In this field it is used to solve the Poisson-Boltzmann equation accurately at the molecular surface. (UNVERIFIED) immersed interface analysis An enhanced sampling method that generates new conformations by displacing a structure along its slowest collective modes rather than at random. Those modes are the directions in which the molecule moves most easily, so displacing along them reaches genuinely different conformations at low energy cost. Random displacement mostly produces strained structures that are immediately rejected. (UNVERIFIED) low-mode conformational search A simulation method that runs several constant pH simulations in parallel, each at a different pH or temperature, and periodically swaps configurations between them. The swaps let a configuration stuck in one protonation state escape by way of conditions where that state is unfavourable. It converges protonation populations far faster than a single constant pH run. (UNVERIFIED) constant pH replica-exchange molecular dynamics NMR analysis in a simulation context refers to the methods used to calculate NMR-related properties from a molecular dynamics trajectory. This allows for a direct comparison between the simulation results and experimental data from Nuclear Magnetic Resonance (NMR) spectroscopy. This includes calculating observables like J-couplings, nuclear Overhauser effects (NOEs), and relaxation times, which are crucial for validating the accuracy of the simulation. NMR analysis The r-6 weighted distance averaging is a specific calculation method used in NMR analysis to compute an effective distance from a simulation trajectory that can be directly compared to an experimental nuclear Overhauser effect (NOE). Because the NOE intensity is proportional to the inverse sixth power of the distance between two protons, a simple average of the distance is incorrect. For experts, performing this <r⁻⁶>⁻¹/⁶ averaging is the physically correct way to back-calculate NOE-derived distances from a dynamic ensemble of structures. r-6 weighted distance averaging analysis This is an enhanced sampling method that combines the replica-exchange method with self-guided Langevin dynamics (SGLD). It runs multiple SGLD simulations in parallel, typically at different temperatures, and attempts to swap their coordinates. This combination aims to leverage the accelerated local exploration of SGLD with the global barrier-crossing capability of replica-exchange for very efficient conformational sampling. replica-exchange self-guided Langevin dynamics A modelling method that derives the numerical parameters of a force field by fitting its energy functions to reproduce reference data from quantum mechanical calculations or experimental measurements. The parameterization determines how accurate the resulting force field is and for which molecules it can be trusted. (UNVERIFIED) force field parameterization modeling A parameterization method that derives bond and angle force constants from the matrix of second derivatives of the energy (the Hessian) from a quantum mechanical calculation. For each pair of atoms it takes the 3-by-3 block of that matrix that couples them and uses its eigenvalues and eigenvectors, projected onto the bond or angle, to obtain a force constant without any fitting. It is often used to obtain bonded parameters for metal sites and other molecules that existing force fields do not cover. (UNVERIFIED) Seminario method modeling Temperature replica-exchange molecular dynamics is the standard and most common form of the REMD method. It involves running multiple parallel simulations at a range of different temperatures and attempting to swap their coordinates. This is a robust and widely used method for enhancing the sampling of systems with complex energy landscapes, such as in protein folding simulations. temperature replica-exchange molecular dynamics Well-Tempered Adaptively Biased Molecular Dynamics (WT-ABMD) is an enhanced sampling technique in computational chemistry. It merges the features of adaptively biased molecular dynamics (ABMD) and well-tempered metadynamics to calculate the free energy surface of complex molecular systems while avoiding over-exploration of high-energy states. well-tempered adaptively biased molecular dynamics A parameterization method in which the molecule's geometry is written in internal coordinates, as a Z-matrix of bond lengths, bond angles and dihedral angles, and parameters are fitted in those terms. This makes it easy to scan 1 degree of freedom at a time, such as a dihedral angle, and to fit the parameter that controls it. (UNVERIFIED) Z-matrix parameterization modeling A density fitting method that expands the electron density in spherical harmonic functions centred on each atom, so that electron repulsion integrals can be computed more cheaply. Fitting the density rather than computing every integral exactly is what makes large density functional calculations affordable. The error it introduces is small and controllable. (UNVERIFIED) ZLM density fitting A force interpolation method is a numerical technique used in some simulation algorithms to estimate the forces on atoms without calculating them directly at every step. This is often done by calculating the forces on a coarser grid and then interpolating these forces to the actual atomic positions. For experts, this method can be used in some multiple-time-stepping algorithms to reduce the computational cost of the force calculation. (UNVERIFIED) force interpolation analysis A thermostat algorithm that adapts its removal of heat during the simulation so that it can absorb extra heat produced by forces that do not conserve energy, such as the mixed forces of adaptive QM/MM simulations. Examples are the adaptive Langevin, adaptive Nosé-Hoover chain and adaptive Nosé-Hoover (chain)-Langevin thermostats. (UNVERIFIED) ntt=6 ntt=7 ntt=8 adaptive thermostat ntt=6 ntt=7 ntt=8 An algorithm that governs how a molecular simulation advances in time and how its physical conditions are held steady. It covers the integrator that moves the atoms forward in time, the methods that compute the forces and share the work between processors, and the thermostats and barostats that keep temperature and pressure at their targets. The integrator, thermostat and barostat together determine which ensemble, that is which set of fixed conditions such as constant temperature, the simulation samples. (UNVERIFIED) simulation control algorithm A numerical analysis that applies general mathematical procedures, such as fitting, transforming, integrating and differentiating, to the output of a simulation in order to extract quantities from it. It is distinguished from the domain-specific analyses by making no assumption about what the numbers represent. (UNVERIFIED) mathematical analysis An algorithm that computes the force on every atom at each step, from the positions of the atoms and the force field. It dominates the cost of a simulation, because the non-bonded part must consider a great many pairs. Almost every optimisation in a simulation engine is aimed at this step. (UNVERIFIED) force calculation algorithm An Ewald summation method in which the charge distribution is represented by charges assigned to atoms by Mulliken population analysis rather than by fixed force field charges. It lets the charge distribution from a quantum mechanical calculation be treated with the standard methods for periodic electrostatics. Its accuracy is limited by how well charges placed on the atoms represent the real distribution of electrons. (UNVERIFIED) Mulliken Ewald summation A numerical method for computing energy derivatives in a calculation with more than one quantum layer, such as a high-accuracy method inside a cheaper one. The derivatives of each layer must be combined so that the force on a shared atom is consistent. Getting the combination wrong produces forces that do not match the energy, and a simulation that drifts. (UNVERIFIED) QM-QM derivative calculation analysis An application of Ewald summation to a system whose charge distribution comes from a quantum mechanical calculation rather than from fixed charges. It is needed because a quantum region in a periodic box interacts with its own images, which must be accounted for rather than ignored. The difficulty is that the charge distribution changes as the calculation proceeds. (UNVERIFIED) QM Ewald treatment A multiscale simulation in which a small chemically active region is described by quantum mechanics and the rest by a force field. It is what allows a reaction inside an enzyme to be modelled, since a force field cannot break bonds and a quantum treatment of the whole protein is far too expensive for most purposes. The difficulty is always the boundary, where the 2 descriptions must be joined without artefacts. (UNVERIFIED) QM/MM QM/MM simulation A QM/MM method in which atoms in a buffer zone between the quantum and classical regions receive a smoothed mixture of the forces from a larger and a smaller quantum calculation. This lets atoms cross the quantum boundary during the run without sudden jumps in force. It was first applied to an ion dissolved in a solvent, where solvent molecules move in and out of the quantum region around the ion. (UNVERIFIED) hot spot adaptive QM/MM An electrostatic interaction algorithm that estimates long-range interactions by treating the region within a cutoff around each particle as repeating in the same way in all directions, and summing the interactions with these imagined images. Unlike Ewald summation, which sums over the actual periodic images of the simulation box, it uses images of a local region, so it works best when the system has a similar composition throughout; a variant using fast Fourier transforms handles non-uniform systems. It can be applied to electrostatic interactions, Lennard-Jones interactions or both. (UNVERIFIED) isotropic periodic sum method A semi-isotropic position scaling barostat is a pressure control algorithm that couples the scaling of the x and y dimensions of the simulation box together, while allowing the z dimension to fluctuate independently. This method is specifically designed for simulating planar systems like lipid bilayers, where the pressure in the plane of the membrane is different from the pressure perpendicular to it. semi-isotropic pressure scaling barostat semi-isotropic position scaling barostat An electrostatic method that computes the force between every pair of charged atoms explicitly and adds them all up. It is exact for a system with no periodic images, and its cost grows with the square of the number of atoms. That growth is why it is used only for small or non-periodic systems. (UNVERIFIED) direct sum Coulomb method Self-guided Langevin dynamics (SGLD) is an enhanced sampling method that accelerates conformational searching by adding a memory-dependent guiding force to the standard equations of motion. This guiding force is calculated from a local average of the forces the system has recently experienced, effectively pushing the system to explore new regions faster. SGLD enhances sampling by reducing random fluctuations and promoting more directed exploration of the potential energy surface. self-guided Langevin dynamics A self-guided Langevin dynamics method in which the random force is modified, following a generalized Langevin equation, so that the guided motion still samples the correct equilibrium distribution at constant temperature or constant pressure. Like the original method, it speeds up slow, low-frequency motions so that conformations are explored faster. Unlike the original, its results need no reweighting afterwards. (UNVERIFIED) generalized Langevin equation self-guided Langevin dynamics Accelerated molecular dynamics is a special type of simulation whose scope is to overcome the timescale limitations of classical molecular dynamics. It works by adding a "boost" to the system's energy, effectively lowering the energy barriers that separate different molecular conformations. For experts, aMD is an enhanced sampling method that adds a non-negative, history-independent bias potential when the system's potential energy is below a certain threshold, allowing for the recovery of canonical averages through energetic reweighting. aMD accelerated molecular dynamics Gaussian accelerated molecular dynamics (GaMD) is an enhanced sampling method that speeds up simulations by adding a boost potential that has a Gaussian distribution. This method smooths the potential energy surface, allowing the system to more easily escape from local energy minima. GaMD Gaussian accelerated molecular dynamics Ligand Gaussian accelerated molecular dynamics is a specific application of the GaMD method focused on enhancing the sampling of a ligand's movement and conformational changes. The boost potential is applied in a way that specifically accelerates the dynamics of the ligand, its binding site, or both. ligand Gaussian accelerated molecular dynamics Dual-boost ligand Gaussian accelerated molecular dynamics (Dual-boost LiGaMD) is a variant of the GaMD method where two separate boost potentials are applied simultaneously to enhance the sampling of ligand binding and unbinding pathways. For example, one boost is applied to the dihedral angles of the ligand to enhance its conformational sampling, while a second boost is applied to the total potential energy of the system to accelerate its overall dynamics. dual-boost ligand Gaussian accelerated molecular dynamics Peptide Gaussian accelerated molecular dynamics is a specific application of the GaMD method tailored to enhance the conformational sampling of peptides. The boost potential is applied to accelerate the slow folding and conformational transition events that are characteristic of flexible peptide molecules. peptide Gaussian accelerated molecular dynamics A method that uses probability theory to collect, summarise and draw conclusions from data. It is what puts an uncertainty on a value and what decides whether a difference between 2 results is real. Simulation output is a sample, so nothing taken from it is complete without one. (UNVERIFIED) statistical analysis Dual-boost peptide Gaussian accelerated molecular dynamics (Dual-boost PepGaMD) is a variant of GaMD where two separate boost potentials are applied simultaneously to enhance the sampling of peptides. One boost is typically applied to the peptide's dihedral angles to accelerate local conformational changes, while the second is applied to the total system potential to overcome larger energy barriers. dual-boost peptide Gaussian accelerated molecular dynamics Ligand Gaussian accelerated molecular dynamics 3 (GaMD3) is a recent and refined version of an enhanced sampling method used to accelerate the simulation of ligand binding and unbinding events. It applies a smoothly varying boost potential to the system's energy landscape, allowing the ligand to escape from deep binding pockets and explore different conformations more rapidly. This third-generation implementation includes improved algorithms for updating the boost potential, aiming for more robust and efficient sampling of complex ligand dynamics. ligand Gaussian accelerated molecular dynamics 3 This is an application of the GaMD method specifically designed to enhance the sampling of protein-protein binding and unbinding events. The boost potential is applied to the potential energy of the entire complex to accelerate the slow association and dissociation processes. protein-protein interaction Gaussian accelerated molecular dynamics Dual-boost protein-protein interaction Gaussian accelerated molecular dynamics (Dual-boost PPI-GaMD) is a variant of GaMD where two separate boost potentials are applied simultaneously to enhance the sampling of protein-protein interfaces. One boost is typically applied to the protein protein interface to accelerate local conformational changes, while the second is applied to the total system potential to overcome larger energy barriers. dual-boost protein-protein interaction Gaussian accelerated molecular dynamics Multiply-targeted molecular dynamics is an enhanced sampling technique that biases a simulation to explore the conformational space between several known target structures simultaneously. Unlike standard targeted MD which pulls a molecule from one state to another, this method encourages the system to visit a collection of different functional states. This is useful for characterizing the pathways that connect multiple known conformations of a protein, such as its active and inactive forms. multiply-targeted molecular dynamics A Partial Nudged Elastic Band (PNEB) is a computational method used to calculate the minimum energy path for structural transitions. Unlike traditional NEB, it applies the nudging algorithm only to a user-defined subset of atoms while allowing the remaining atoms to optimize freely. This reduces computational costs and supports complex environments like explicit solvents. partial nudged elastic band An algorithm that searches for the input values that give the lowest or highest value of a given function. In molecular simulation the function is often an energy and the inputs are atomic positions, as in energy minimization, but it can also be the error of a fit to data or the energy of an electronic structure calculation. Many such algorithms find only the nearest minimum, not the lowest one overall. (UNVERIFIED) optimization algorithm An optimization algorithm that adjusts atomic positions to lower the energy of a structure until it reaches a nearby minimum. It moves downhill only, so it settles in the nearest low point rather than the lowest one. It is run on a freshly built system to remove atomic overlaps, whose very large forces would otherwise make the first simulation steps fail. (UNVERIFIED) energy minimization algorithm An energy minimization algorithm package in AMBER that offers 3 local minimization methods: Polak-Ribière conjugate gradient, limited-memory BFGS, which is the default, and truncated Newton conjugate gradient. It is also used for the energy relaxation steps of the LMOD conformational search method. (UNVERIFIED) XMIN minimization algorithm A free energy analysis method in which the alchemical coupling parameter, lambda, is replaced by a smooth function of lambda that can be chosen separately for each kind of interaction, such as electrostatic, van der Waals or bonded terms. Each interaction can then be switched over its own part of the lambda range, for example van der Waals terms only between lambda values of 0.5 and 1. This gives finer control over how the 2 end states are mixed in a thermodynamic integration calculation. (UNVERIFIED) lambda-scheduling free energy analysis Hybrid solvent replica-exchange molecular dynamics is a specialized version of the REMD method where the different parallel simulations use different solvent models. For example, some replicas might use an implicit solvent model, which allows for very fast sampling, while others use a more accurate explicit solvent model. Swapping between these replicas allows the simulation to combine the sampling efficiency of implicit solvent with the accuracy of explicit solvent. hybrid solvent replica-exchange molecular dynamics Reservoir replica-exchange molecular dynamics is an efficient variant of the REMD method for simulations with a very large number of replicas. Instead of attempting swaps between all adjacent pairs, it attempts swaps between a master simulation and a large "reservoir" of simulations running at different temperatures. This can significantly improve the efficiency of the exchange process when a very wide temperature range needs to be sampled. reservoir replica-exchange molecular dynamics Replica-exchange umbrella sampling is an enhanced sampling method that combines two powerful techniques to calculate free energy profiles. It runs multiple umbrella sampling simulations in parallel, each restraining the system to a different window along a reaction coordinate, and periodically attempts to swap the coordinates between these simulations. This combination helps to overcome slow conformational sampling within each umbrella window, leading to faster convergence of the potential of mean force. replica-exchange umbrella sampling Replica-exchange with an arbitrary degree of freedom is a powerful generalization of the standard replica-exchange simulation method. Instead of running parallel simulations at different temperatures, this technique allows them to be run along any user-defined variable, such as a reaction coordinate, a force field parameter, or a pH value. By attempting to swap these variables between replicas, the method can dramatically enhance sampling and map out a system's response to a wide range of conditions in a single set of simulations. replica-exchange with arbitrary degree of freedom A free energy calculation analysis that computes an absolute free energy by thermodynamic integration from a realistic atomistic model to a simplified reference model whose free energy can be solved exactly. The Einstein Molecule model is used for the solute, in which each atom is held in place by a harmonic potential so that the reference free energy has a closed-form solution. EMIL is the name of the module that implements this in AMBER, and its authors call the underlying method Thermodynamic Integration relative to an Exact Solution. (UNVERIFIED) Berryman J.T. Absolute free energies for biomolecules in implicit or explicit solvent. Physics Procedia, 2014. States that the AMBER implementation of this method is called EMIL. Berryman J.T. and Schilling T. Free energies by thermodynamic integration relative to an exact solution, used to find the handedness-switching salt concentration for DNA. J. Chem. Theory Comput. 9(1):679-686, 2013. EMIL free energy analysis A simulation method in which acidic and basic groups can gain or lose a proton during the run, rather than being fixed at the start. Their protonation is coupled to a chosen acidity, so the populations that result are those at that pH. It matters because a fixed assignment made before the simulation may be wrong, and a wrong charge changes everything downstream. (UNVERIFIED) CpHMD constant pH molecular dynamics A constant pH method in which the degree of protonation of each titratable group is a continuous quantity that moves under its own equation of motion, rather than switching between protonated and deprotonated. Because it changes smoothly, no jump has to be accepted or rejected and the simulation never stops to attempt one. The cost is that intermediate values do not correspond to any real chemical state and must be kept away from. (UNVERIFIED) continuous constant pH molecular dynamics Lambda-dynamics is a free energy calculation method where the coupling parameter, lambda, is treated as a dynamic variable that evolves in time, just like an atomic coordinate. A fictitious mass is assigned to lambda, and its "motion" along the transformation pathway is simulated, allowing it to sample different intermediate states. lambda-dynamics free energy analysis A continuous constant pH method in which both the motion of the atoms and the changes in protonation are computed in explicit solvent, with the long-range electrostatic interactions calculated by particle mesh Ewald. In the other continuous variants, the forces that drive protonation changes come from an implicit (generalized Born) solvent model. (UNVERIFIED) PME continuous constant pH molecular dynamics A free energy analysis method that builds a free energy profile from biased simulations, such as umbrella sampling, by finding the profile that makes the observed samples most probable, a maximum likelihood approach. It removes the effect of the biasing potentials, and AMBER offers it alongside the multistate Bennett acceptance ratio for this purpose. (UNVERIFIED) variational free energy profile analysis A free energy analysis method that obtains the free energy profile for an expensive target potential, such as ab initio QM/MM, from biased simulations run with a cheaper reference potential, such as semi-empirical QM/MM, by reweighting the sampled configurations. Only the energies of the sampled configurations need to be computed with the expensive potential, not a full simulation. (UNVERIFIED) weighted thermodynamic perturbation theory analysis A free energy analysis method that estimates the free energy profile of an expensive high-level potential, such as an ab initio QM/MM model, from umbrella sampling simulations run with several cheaper low-level reference potentials. It generalises weighted thermodynamic perturbation, which uses a single reference potential, to many reference potentials. The cheaper potentials allow far more sampling than the high-level potential would. (UNVERIFIED) generalized weighted thermodynamic perturbation theory Multidimensional replica-exchange molecular dynamics is an extension of the REMD method where exchanges are performed along more than one variable simultaneously. For example, parallel simulations could be run at different temperatures and different pH values, with swaps attempted in both the temperature and pH dimensions. multidimensional replica-exchange molecular dynamics An optimization algorithm that locates a first-order saddle point on a potential energy surface, where the energy is at a maximum along 1 direction and a minimum along all others. It differs from a minimiser in having to climb in a chosen direction rather than descend in every one. The saddle point it finds is the transition state that sets a reaction barrier. (UNVERIFIED) transition state search algorithm A transition state search algorithm that uses 2 closely spaced copies of a structure to find the direction in which the energy curves downwards most, then moves uphill along it to a saddle point. It needs only first derivatives of the energy, never the full matrix of second derivatives. That makes it usable on systems too large for methods that build the whole Hessian. (UNVERIFIED) dimer method algorithm A data analysis method that computes the local density, energy and entropy of water on a 3-dimensional grid around a solute, from a simulation with explicit solvent. It assigns thermodynamic quantities to positions rather than to the system as a whole. That is what allows individual hydration sites to be identified and their contribution to solvation free energy to be quantified. (UNVERIFIED) grid inhomogeneous solvation theory analysis An analysis method that builds a 3-dimensional map showing which regions around a molecule matter for its function, by computing a chosen property at each point of a grid. What the map shows depends on the property: where a probe binds favourably, where solvent is held, or where charge accumulates. It turns a per-point calculation into a picture that can be inspected. (UNVERIFIED) 3D-mapping analysis An electrostatic method that computes the total electrostatic energy of a periodically repeating system by splitting the calculation into a short-range part summed directly in real space and a long-range part summed in reciprocal (Fourier) space. The split works because each part converges quickly where the other does not. Summed directly over the periodic images, the Coulomb sum converges only slowly and conditionally, meaning its value depends on the order in which the terms are added. (UNVERIFIED) Ewald summation An algorithm that computes the long-range part of the Lennard-Jones dispersion interaction, the attraction that falls off with the sixth power of distance, using the same grid-based Ewald treatment that particle mesh Ewald applies to electrostatics. Dispersion is usually cut off at a fixed distance, sometimes with an average correction that assumes the system is uniform beyond the cutoff, which causes errors in non-uniform systems such as membranes and interfaces. Treating the long-range part explicitly reduces these errors at a modest extra cost. (UNVERIFIED) Lennard-Jones particle mesh Ewald The part of a force field describing interactions between atoms not joined by a bond, comprising the electrostatic term and the van der Waals term. Every atom in principle interacts with every other, so this part dominates the cost of a simulation. Nearly all the effort of making simulations fast goes into computing it. (UNVERIFIED) non-bonded potential A force field component that keeps a group of atoms flat, or keeps the correct handedness at a chiral centre, by penalising movement of a central atom out of the plane of the 3 atoms bonded to it. The 4 atoms involved are not a chain of bonded atoms, which is why the term is called improper. Some force fields write it as a harmonic term with a single preferred value, while others, such as AMBER, use a repeating cosine term like a proper dihedral. (UNVERIFIED) improper torsion potential A set of mathematical functions, often resembling atomic orbitals, used in quantum chemical calculations to represent the spatial distribution of electrons within a molecule (molecular orbitals). The choice and size of the basis set significantly impact the accuracy and computational cost of a simulation; larger, more flexible basis sets generally yield more accurate results but require more computer resources. These functions serve as the building blocks for constructing the calculated electronic wavefunction. (UNVERIFIED) quantum mechanics basis set A structural analysis method that describes the shape of a 5-membered ring, such as the ribose sugar of a nucleotide, using 2 numbers: the phase angle of pseudorotation and the puckering amplitude. The phase angle says which atoms lie out of the plane and in which direction, and the amplitude says by how much. Together they place a sugar in a named conformational family such as C2'-endo or C3'-endo. (UNVERIFIED) Altona & Sundaralingam pucker analysis A structural analysis method that describes the shape of any non-planar ring using a set of numbers derived from how far each atom sits out of the ring's mean plane. It applies to rings of any size, where the older 2-parameter treatment covers 5-membered rings only. For a 6-membered ring it gives 3 numbers that place it on a sphere of possible shapes. (UNVERIFIED) Cremer & Pople pucker analysis A simulation method that adds a correction, learned by a kernel machine, to the energies and forces of a classical force field so that they approach those of a more accurate reference method, such as ab initio quantum mechanics. The kernel machine is trained on configurations labelled with both energies and predicts the difference for a new configuration from its similarity to the training configurations. In the AMBER implementation the similarity is measured on chosen dihedral angles. (UNVERIFIED) kernel modified molecular dynamics A defined set of rules and conventions for how data are laid out, structured and encoded, most often within a computer file. In molecular simulation, separate formats exist for structures, topologies, trajectories, input settings and results, and a program can read only the formats it was written to handle. (UNVERIFIED) data format A data format that provides the necessary data, settings, or instructions required to initiate a computational task. This format provides the instructional input for simulation or analysis software. It is distinct from formats that describe molecular structure or results. (UNVERIFIED) simulation input format A data format for a file containing a sequence of commands or code in a specific language. These scripts are designed to be executed by a program to automate a workflow or a series of analysis steps. script format A simulation log format is a text-based file structure that records the progress and key information generated during a computer simulation or calculation. It typically includes a copy of the input settings, step-by-step updates on energy and temperature, performance timings, and any warnings or error messages. This format is the primary source for monitoring a simulation's stability, diagnosing problems, and extracting averaged thermodynamic properties after the run is complete. output log format simulation log format A category of data formats for files containing information derived from Nuclear Magnetic Resonance (NMR) experiments. These formats are used to store and exchange experimental data, such as structural restraints, that can be used in simulations. NMR data format A data format for a file that stores a complete snapshot of a simulation's state at a specific time. This includes coordinates, velocities, and other state variables sufficient to resume the run. These formats are often binary for efficiency. simulation state format A text data format that stores the results of a quantum chemistry calculation for visualization, holding atomic coordinates together with the basis functions and coefficients that describe the molecular orbitals. Because it carries the orbitals themselves rather than only a picture of them, a viewer can render orbitals, electron density and vibrational modes from it. It is read by many visualization programs, not only the one it is named after. (UNVERIFIED) Molden format A text data format for volumetric data in which each line gives the x, y and z coordinates of 1 grid point followed by the value at that point, with no header or footer. Because every point carries its own coordinates, it is simple to read but larger and slower than formats that state the grid once. AMBER 3D-RISM calculations can write their solvent distributions in it. (UNVERIFIED) xyzv format A script file format for the tleap program in AmberTools, used to build a molecular system for simulation. It contains a sequence of commands to load molecular structures, apply force fields, add solvent, and save the final topology and coordinate files. This format is a fundamental part of the system preparation stage in an Amber workflow. (UNVERIFIED) tleap input format The mmCIF format (.cif or .mmcif) is the standard archival format for macromolecular structures used by the Protein Data Bank (PDB), superseding the legacy PDB format. It uses a flexible, dictionary-based syntax (CIF) with key-value pairs and loop structures to store atomic coordinates, connectivity, experimental details, and extensive metadata. It is better suited for handling large, complex structures and diverse experimental data. mmCIF format The PDB format (.pdb) is the long-standing, fixed-column text format traditionally used for storing protein and nucleic acid structures from the Protein Data Bank. It includes records for metadata, atomic coordinates, and optionally connectivity. While still widely compatible with visualization and simulation software, it is considered obsolete since it has limitations for very large structures and complex metadata compared to the newer mmCIF format. PDB format A name, code or label assigned to an entity, a data record or a concept that distinguishes it unambiguously from others within a given context. Accession numbers in databases, chain letters in a protein structure and atom names in a topology are all examples. Identifiers are what let a piece of information be referred to, retrieved and linked reliably. (UNVERIFIED) identifier OBSOLETE. A protein database ID is a unique alphanumeric code assigned to each entry in a public protein structure or sequence database, such as the Protein Data Bank (PDB ID, e.g., '1TIM') or UniProt (accession number, e.g., 'P00761'). These identifiers serve as the standard way to reference and retrieve specific protein structures or sequence information. Retired 2026-08-07 on review by orcid:0000-0002-2294-5393. Two objections, both accepted. First, the Protein Data Bank is not a protein database: it holds nucleic acids, protein-nucleic acid complexes and other assemblies, so the label and definition made a claim that is simply false. Second, a PDB entry identifies a deposited structure while a UniProt accession identifies a sequence record, so grouping them said less than it appeared to. Its two children, MOLSIM:000686 PDB ID and MOLSIM:000687 UniProt ID, were re-parented to MOLSIM:000447 identifier alongside the other identifiers. No IAO:0100001 successor, because there is no single term that replaces it; this is the same shape as the other retired grouping terms whose children were re-parented instead. Note the term had been verified by orcid:0000-0003-4177-3619 before this review. obsolete protein database ID true A function that reduces the electrostatic interaction between 2 charges as they approach very closely. Without it, a polarizable model can reach a state where 2 induced dipoles reinforce each other without limit and the energy diverges. Damping prevents that failure, which is called polarization catastrophe. (UNVERIFIED) tinker-exponential damping electrostatic damping function Hydrogen bond analysis is a structural analysis method used to identify and quantify the hydrogen bonds that are formed and broken during a simulation. The analysis typically uses a set of geometric criteria, such as a maximum distance between the donor and acceptor atoms and a minimum angle, to define when a hydrogen bond exists. hydrogen bond analysis A structural analysis method that tests whether the orientations of the structures in a collection are spread evenly over all directions. It computes the distribution of orientations and looks for any preferred alignment. It is used to confirm that a simulation has not introduced an artificial bias in rotational motion, which some theoretical treatments assume is absent. (UNVERIFIED) isotropically distributed ensemble analysis A dynamics analysis method that decomposes the reorientational motion of a set of bond vectors in a molecule, such as backbone N-H bonds, into collective modes, by diagonalising a matrix built from the relative orientations of all pairs of these vectors. It assumes that the overall tumbling of the molecule is isotropic, meaning the same in all directions. Each mode has its own correlation time, and together they give NMR order parameters and relaxation rates that can be compared with experiment. (UNVERIFIED) isotropic reorientational eigenmode dynamics analysis Nucleic acid structure analysis is a specialized type of structural analysis that calculates a set of parameters describing the detailed geometry of a DNA or RNA helix, known as the helical parameters. This includes parameters that describe the geometry of individual base pairs as well as the relationship between adjacent base pairs, such as twist, roll, and slide. nucleic acid structure analysis A structural analysis method that tracks how many of the atom pairs in contact in a reference folded structure are also in contact at each point in a trajectory. The reference set is taken from the native state and the result is reported as the fraction of it that is present. That fraction is a common single-number measure of how far folding has progressed. (UNVERIFIED) native contact analysis A statistical method that finds the directions in which a system changes most slowly, rather than those in which it varies most. It does this by looking at how coordinates relate to themselves after a delay rather than at the same instant. The slowest motions are usually the functionally important ones, and they are not always the largest. (UNVERIFIED) tICA time-lagged independent component analysis A method that decomposes a signal into contributions at different frequencies, while keeping track of when each occurred. A standard frequency analysis says which frequencies are present over the whole record but not when. That matters for a trajectory, where a motion may appear only during part of the run. (UNVERIFIED) wavelet analysis A wavelet analysis that groups the significant points of a wavelet map of a trajectory, using a density-based clustering algorithm, into regions of atoms and frames that show significant motion. It picks out which atoms move, and when, rather than describing the trajectory as a whole. The atoms and frames in each region can then be extracted for further analysis. (UNVERIFIED) wavelet analysis feature extraction Fluorescence spectroscopy is an experimental technique that measures the light emitted by a substance after it has absorbed light. A molecule absorbs light at one wavelength, promoting it to an excited electronic state, and then emits light at a longer, lower-energy wavelength as it relaxes back down. This technique is highly sensitive and is widely used to study molecular conformation, binding, and dynamics by monitoring changes in the fluorescence signal. fluorescence spectroscopy A data format used by GROMACS that defines a reusable template for one residue, giving its atoms, their charges and the bonds between them. It is what lets a topology be built automatically from a sequence rather than written out atom by atom. It is the GROMACS counterpart of the CHARMM residue topology file. (UNVERIFIED) GROMACS rtp format rtp format An endpoint free energy tool that estimates binding or solvation free energies from simulation snapshots by combining molecular mechanics energies, an electrostatic solvation energy from the Poisson-Boltzmann equation and a nonpolar term proportional to the solvent accessible surface area. It is typically used to estimate ligand binding affinities. It is computationally cheaper than free energy perturbation or thermodynamic integration but generally less rigorous. (UNVERIFIED) MMPBSA tool molecular mechanics Poisson-Boltzmann surface area tool A molecular mechanics Poisson-Boltzmann surface area tool that is built into a larger simulation or modelling package rather than run on its own. It uses that package's own energy routines and Poisson-Boltzmann solver, which is what distinguishes it from a standalone script such as MMPBSA.py. (UNVERIFIED) MMPBSA module molecular mechanics Poisson-Boltzmann surface area module An endpoint free energy tool that estimates binding or solvation free energies from simulation snapshots by combining molecular mechanics energies, a generalized Born electrostatic solvation energy and a nonpolar term proportional to the solvent accessible surface area. The generalized Born model approximates the Poisson-Boltzmann equation, so the calculation is faster than the Poisson-Boltzmann variant but often less accurate. Like that variant, it is cheaper but less rigorous than alchemical free energy methods. (UNVERIFIED) MMGBSA tool molecular mechanics generalized Born surface area tool A Tversky index algorithm that adds the shape Tversky score and the color (chemical feature) Tversky score of 2 aligned molecules, both weighted towards the reference (query) molecule with α 0.95 and β 0.05. Considering shape and chemical features together gives a fuller comparison than either alone, and favors database molecules that contain the query. The combined score normally lies between 0 and 2. (UNVERIFIED) ref Tversky combo A minimal basis set in which each Slater-type atomic orbital is approximated by a fixed combination of 6 Gaussian functions. It gives the closest approximation to Slater-type orbitals in the family, but adding Gaussian functions never lets the electron cloud change shape as a bond forms. It therefore shares the limited accuracy of all minimal basis sets. (UNVERIFIED) STO-6G NModelGB is a specific implementation of a Generalized Born (GB) implicit solvent model, likely used within a particular simulation package. It provides a specific set of equations and parameters for calculating the electrostatic contribution to the solvation free energy. The "NModel" designation distinguishes this particular variant from other GB models available to the user. (UNVERIFIED) NModelGB OPM database (Orientations of Proteins in Membranes) is both a database and a web server/tool providing pre-calculated transmembrane orientations and positions for experimentally determined membrane protein structures found in the PDB. OPM uses a computational approach based on minimizing the calculated transfer free energy from water to the lipid bilayer to determine the most likely membrane boundaries and protein position relative to the bilayer center. It's a valuable resource for obtaining initial membrane protein placements for simulations. OPM orientations of proteins in membranes Förster Resonance Energy Transfer (FRET) is a specific phenomenon in fluorescence spectroscopy where an excited donor fluorophore non-radiatively transfers its energy to a nearby acceptor fluorophore. The efficiency of this energy transfer is extremely sensitive to the distance between the two molecules, making it a "spectroscopic ruler." FRET Forster resonance energy transfer A binary file format used to store three-dimensional images or maps, common in electron microscopy. It is the standard format for storing 3D density maps in cryo-electron microscopy (cryo-EM) and is also used in X-ray crystallography. This format is essential for the visualization and analysis of volumetric data in structural biology. (UNVERIFIED) MRC format A text-based format containing control parameters for the Amber sander and pmemd programs, organized in Fortran namelists. It specifies simulation conditions like timestep, duration, and temperature/pressure coupling algorithms. This format dictates how the molecular dynamics simulation is performed. AMBER mdin format A text-based log format generated by Amber's simulation engines that provides a detailed record of the simulation's progress. It contains energy components, thermodynamic properties, and performance timings printed at regular intervals. This format is essential for monitoring simulation stability and extracting averaged properties. (UNVERIFIED) AMBER mdout format AMBER out format mdout format A format that stores the complete state of an Amber simulation, including atomic coordinates, velocities, and periodic box dimensions. It is used to save the simulation state at a specific point in time, allowing the run to be resumed exactly. The format can be either ASCII-based (.rst7) or binary NetCDF-based (.ncrst). AMBER restart format The Amber molecular dynamics topology file (.prmtop or .parm7) contains all the static information about the molecular system needed for an Amber simulation, excluding coordinates. This includes atom definitions (names, types, masses, charges), molecular connectivity (bonds, angles, dihedrals), force field parameters for all interaction terms, and non-bonded interaction rules. It is generated by programs like tleap and used together with a coordinate file (.inpcrd, .rst7). AMBER topology format A file format generated by the Antechamber program in AmberTools that contains a molecule's structure and atomic properties. It typically stores atom types and partial charges assigned during the parameterization process for a small molecule. This format serves as an intermediate in preparing non-standard molecules for simulation. (UNVERIFIED) ac format A computational method that predicts how a small molecule sits in a binding site on a protein, and scores how well it binds. It searches over positions and shapes of the molecule and rates each by a scoring function. It is fast enough to screen large libraries, and its scores rank candidates far better than they predict real binding strengths. (UNVERIFIED) protein-ligand docking analysis A molecular dynamics method that simulates a model in which groups of atoms are merged into single interaction sites, often called beads. Having fewer particles and smoother interactions allows a longer timestep, so larger systems and longer timescales can be reached than with all-atom simulation. The cost is that no detail finer than 1 bead can be seen. (UNVERIFIED) CG-MD coarse-grained molecular dynamics OBSOLETE. A specific, widely used 3-point water model type characterized by its simplicity and computational efficiency (Transferable Intermolecular Potential 3-Point). It places charges on the oxygen and two hydrogen atoms and represents water as a rigid molecule with experimentally derived geometry. While computationally fast, it is known to have limitations in reproducing some properties of real water (like the temperature of maximum density or dielectric constant) compared to more complex models. Retired 2026-08-24 as a duplicate. MOLSIM:000022 TIP3P describes the same rigid three-site model with charges on the oxygen and two hydrogens, and sits with the other 18 water models under MOLSIM:000014 water model force field. This one sat under the parallel MOLSIM:000030 water type branch, which was retired in the same change. obsolete TIP3P water model true A method that assigns charge to atoms by fitting point charges to reproduce the molecule's electrostatic potential at points on a grid outside it. Fitting to the potential is what makes the resulting charges suitable for a force field, since that potential is what other molecules feel. Charges on buried atoms are poorly determined, because the potential outside barely depends on them. (UNVERIFIED) CHELPG charge analysis Molecular surface analysis is a category of analysis methods used to calculate and characterize the surface of a molecule. This includes calculating properties like the total surface area, identifying pockets and cavities, and mapping properties like electrostatic potential onto the surface. molecular surface analysis A Connolly surface calculation is a specific algorithm used to compute the solvent-accessible surface of a molecule, which is the surface traced out by the center of a probe sphere as it rolls over the molecule. It is a widely used method for defining the boundary between a molecule and the surrounding solvent. Connolly surface analysis A dynamics analysis that computes the matrix of covariances between the displacements of all pairs of atoms from their average positions over a simulation. Each element shows how strongly the motion of 1 atom is coupled to the motion of another. Diagonalizing this matrix is the central step of principal component analysis, which finds the dominant collective motions of a protein or other macromolecule. (UNVERIFIED) covariance matrix analysis OBSOLETE. A 2D RMSD analysis is a method where the Root-Mean-Square Deviation (RMSD) is calculated for multiple frames in a simulation trajectory with respect to a reference structure. The reference structure is usually the first frame in the trajectory or an average structure of the whole simulation. The results are typically visualized as a two-dimensional plot, where each point represents the RMSD between one frame and the reference structure. Retired as redundant with 'RMSD analysis' (MOLSIM:000311), which shares the same parent and the same underlying principle. The 2D frame-to-frame form is now covered explicitly by that term's definition. Reported and approved by the term verifier in GitHub issue #45. obsolete 2D RMSD analysis true A lipid SCD order parameter calculation is a specific analysis method used to quantify the orientational order of the carbon-hydrogen bonds in the acyl tails of lipid molecules in a membrane simulation. The SCD parameter measures the motional restriction of these bonds relative to the membrane normal. Calculating these order parameters and comparing them to experimental values from NMR is a primary method for validating the accuracy of a membrane simulation. lipid SCD order parameter calculation analysis An analysis method that computes the 3-dimensional distribution of solvent around a solute, and from it the solvation free energy, by solving the integral equations of the reference interaction site model of liquids on a grid. It gives these without simulating solvent molecules moving, so no sampling problem arises. What it cannot give is any individual solvent molecule's behaviour over time. (UNVERIFIED) 3D reference interaction site analysis A QM/MM method in which solvent molecules are reassigned between the quantum and the classical region as they diffuse, rather than being assigned once at the start. With a fixed assignment, the quantum solvent molecules would drift away from the reacting site over time and be replaced near it by classical ones. Many adaptive methods change the treatment of a molecule gradually across a transition region so that the forces do not jump when it is reassigned. (UNVERIFIED) adaptive solvent QM/MM A basis set that represents each valence orbital with 2 or more functions of different spatial extent, while keeping a single function for each core orbital. The extra valence functions let the electron cloud expand or contract as bonding requires, which a minimal set cannot. Splitting only the valence shell concentrates the added cost where bonding happens. (UNVERIFIED) split-valence split-valence basis set A split-valence Gaussian basis set describing each core orbital with 3 contracted primitives and each valence orbital with 2 functions, built from 2 and 1 primitives. It is small enough for large molecules and crude enough that its geometries are only approximate. It is used for a quick first optimisation before a better basis set is applied. (UNVERIFIED) 3-21G A split-valence Gaussian basis set describing each core orbital with 4 contracted primitives and each valence orbital with 2 functions. It differs from the 3-primitive version only in the description of the core, which barely affects bonding. It is rarely used today. (UNVERIFIED) 4-21G An improper torsion potential in which the energy rises with the square of the departure from a single preferred value. Unlike the repeating cosine form it has one minimum, which is what makes it suitable for holding a group flat. The AMBER force fields use a cosine form for the same purpose, so the 2 conventions are not interchangeable. (UNVERIFIED) harmonic improper torsion potential A potential energy function representing the van der Waals interaction between 2 atoms, with a steep repulsion at short range and a weaker attraction falling off with the sixth power of distance. The attraction has a physical basis in dispersion; the repulsion is chosen for computational convenience rather than derived. It is the form almost every force field uses. (UNVERIFIED) Lennard-Jones potential A software program that prepares, corrects and checks a molecular structure before it is used in a simulation. Its tasks include adding atoms the experimental structure does not contain, choosing between alternative positions where the data was ambiguous, and reporting geometry that cannot be right. Experimental structures nearly always need this, because crystallography rarely sees hydrogens and often misses flexible loops. (UNVERIFIED) structure preparation tool A commercial molecular modeling suite for drug discovery and computational chemistry, developed by Tripos and now maintained by Certara. It offers tools for building and editing structures, conformational analysis, pharmacophore modeling, quantitative structure-activity relationships, docking, scoring and visualization, along with molecular mechanics and dynamics. In biomolecular simulation it is used mainly for structure preparation and modeling, particularly in pharmaceutical research. (UNVERIFIED) SYBYL A visualization and analysis program for protein structures that also performs measurement, superposition and simple modelling. It can build a model of one protein on the structure of a related one, and submit that model for refinement. It is now distributed under the name DeepView. (UNVERIFIED) DeepView Swiss PDB viewer A molecular structure visualization tool that draws molecules and macromolecules in 3 dimensions and is known for high-quality images for publication. It also measures distances and angles, draws surfaces, makes animations and can be driven by Python scripts. An open-source version exists alongside a commercial version distributed by Schrödinger. (UNVERIFIED) Pymol A simulation preparation stage in which the atomic coordinates of a molecular system are adjusted to reach a nearby minimum of the potential energy, removing clashes between atoms and strain from the initial model. It uses energy minimization algorithms, such as steepest descent or conjugate gradient, which reduce the forces on the atoms step by step until a convergence criterion is met. (UNVERIFIED) minimization A set of atomic radii, used with matching partial charges, developed by Sitkoff, Sharp and Honig in 1994 so that Poisson-Boltzmann calculations reproduce measured solvation energies of small molecules. In a Poisson-Boltzmann calculation these radii define the boundary between the low-dielectric solute and the high-dielectric solvent. They differ from standard van der Waals radii such as Bondi's, because they were optimized to absorb other energetic effects within that model. (UNVERIFIED) parse radii A Python library that implements the multi-state Bennett acceptance ratio estimator for combining data from simulations at several thermodynamic states. It returns free energy differences between any pair of those states together with equilibrium averages. It is the reference implementation of the estimator and is called by other free energy analysis tools. (UNVERIFIED) pyMBAR A statistical metric used to judge how well a computational method performs, rather than to describe the system being studied. Error measures, correlation measures and ranking measures all belong here. They are what a method is validated against before its predictions are relied on. (UNVERIFIED) computational method statistical measure analysis A statistical metric that scores how well a model ranks a set of compounds, by checking for each pair whether the predicted order matches the measured order and weighting the pair by the difference between their measured values. A wrong order for 2 compounds with very different measured values therefore counts more than a wrong order for 2 similar ones. It is used where a model is meant to guide a choice between compounds rather than to reproduce a number. (UNVERIFIED) predictive index of Pearman analysis A data analysis method that measures how closely predicted relative binding free energies match experimental ones for the same set of ligand pairs. It reports error measures such as root mean square error and mean absolute error, and agreement measures such as the Pearson correlation and the Kendall rank correlation. The rank measures matter most when the purpose is to order ligands rather than to predict a number. (UNVERIFIED) accuracy of ddG prediction metrics measurements analysis A QM/MM scheme in which the quantum region is computed in isolation and its interaction with the surroundings is treated entirely by the force field. The quantum region therefore never feels the charges around it, so its electron distribution does not respond to its environment. It is the simplest scheme and the least accurate where the environment is polar. (UNVERIFIED) mechanical embedding QM/MM A split-valence Gaussian basis set describing each core orbital with 4 contracted primitives and each valence orbital with 2 functions built from 3 and 1 primitives. It was an early step towards the widely used 6-31G family and has largely been replaced by it. (UNVERIFIED) 4-31G A statistical mathematical technique used to assess whether an observed linear relationship (correlation) between 2 variables, calculated from simulation data, is statistically meaningful. It evaluates the probability that such a correlation could have occurred simply by random chance if no true underlying relationship existed. This helps researchers confidently interpret potential connections found in simulation results. (UNVERIFIED) linear correlation significance determination analysis A common specific method for linear correlation significance determination (and statistical testing in general) that involves calculating a probability value, the p-value. The p-value represents the probability of observing the calculated correlation, or a stronger one, assuming that there is actually no real correlation (the null hypothesis). A small p-value (typically below a threshold like 0.05) leads to the conclusion that the observed correlation is statistically significant and unlikely due to chance. (UNVERIFIED) determination by p-value analysis A split-valence Gaussian basis set describing each core orbital with 6 contracted primitives and each valence orbital with 2 functions. The larger core description makes it more accurate than its 3-primitive and 4-primitive relatives for core-dependent properties and no better for bonding. It is rarely used today. (UNVERIFIED) 6-21G A split-valence Gaussian basis set that describes each core orbital with 6 contracted primitives and each valence orbital with 2 functions, an inner one contracted from 3 primitives and an outer one from a single primitive. It is one of the most widely used basis sets in quantum chemistry, because it is far more accurate than a minimal basis set while staying cheap enough for routine work. It is also the starting point most polarization and diffuse variants are built on. (UNVERIFIED) 6-31G A multiscale method that describes a solute quantum mechanically while representing the surrounding solvent with a generalized Born continuum model. The quantum region can then respond electronically to the solvent without any solvent molecules being simulated. It is used where reactivity matters but explicit solvent would be too expensive. (UNVERIFIED) QM generalized Born treatment A time-averaged restraint simulation is an advanced simulation method where the restraining potential is applied not to the instantaneous structure, but to the average of a an experimental one over a period of time. This is particularly useful for incorporating experimental data, like from NMR, which is also an average over time and a molecular ensemble. For experts, this method provides a more physically meaningful way to enforce experimental restraints that reflect the dynamic nature of the molecule. time-averaged restraint simulation A split-valence Gaussian basis set that describes each core orbital with 6 contracted primitives and each valence orbital with 3 functions of different spatial extent, contracted from 3, 1 and 1 primitives. The extra valence function gives more freedom than a double-split set, so it is more accurate for the same molecule. It is usually used together with polarization and diffuse functions rather than on its own. (UNVERIFIED) 6-311G A set of effective core potentials and matching valence basis sets developed at Los Alamos National Laboratory for calculations on heavier elements. The potential stands in for the chemically inert inner electrons, which cuts cost and includes some relativistic effects, while the valence basis describes the electrons that form bonds. Its valence basis sets are revised versions of the older LANL2DZ ones and are used with the same potentials. (UNVERIFIED) LANL08 A set of effective core potentials and valence basis sets from the Los Alamos LANL08 family, with f type polarization functions added on the heavy atoms the potential covers. The f functions give more freedom in describing electron distribution, which matters for bonding in molecules containing heavier elements. (UNVERIFIED) LANL08(f) A Los Alamos effective core potential paired with a valence basis contracted to 6 s-type, 4 p-type, 4 d-type and 2 f-type functions. The potential replaces the inner electrons, and the large valence contraction gives far more flexibility than the standard double-zeta basis of the same family. It is used for heavier elements where the standard basis is too small to describe bonding accurately. (UNVERIFIED) LANL2-[6s4p4d2f] A Los Alamos effective core potential paired with a double-zeta valence basis, for elements from the third row onwards. The potential stands in for the inner electrons, which cuts cost and includes some relativistic effects. It is one of the most widely used combinations of effective core potential and basis set. (UNVERIFIED) LANL2DZ ECP A Los Alamos effective core potential with a double-zeta valence basis, carrying 1 extra d type and 1 extra f type polarization function on the heavy atoms the potential covers. The higher angular momentum functions give the valence basis more freedom, which improves the description of bonding for heavier elements. (UNVERIFIED) LANL2DZ+1d1f A Los Alamos effective core potential with a double-zeta valence basis, carrying 2 additional s, p, d and f functions on the heavy atoms the potential covers. The additions give far more freedom in describing valence electrons than the standard double-zeta set. (UNVERIFIED) LANL2DZ+2s2p2d2f A Los Alamos effective core potential with a double-zeta valence basis, to which diffuse and polarization functions have been added for the p-block elements of groups 14 to 17. The added functions improve the description of anions and of bonding over the unaugmented version, with results for vibrational frequencies and bond energies similar to an all-electron basis of comparable size. (UNVERIFIED) LANL2DZdp ECP A Los Alamos effective core potential with a double-zeta valence basis, carrying d type polarization functions on the heavy atoms the potential covers and p type polarization functions on hydrogen. The added polarization functions improve molecular geometries and interaction energies markedly over the unpolarized version. (UNVERIFIED) LANL2DZdp ECP Polarization A Los Alamos effective core potential paired with a triple-zeta valence basis set, using 3 functions per valence orbital. The extra valence function makes it more flexible and more accurate than the double-zeta member of the same family, at a higher cost. (UNVERIFIED) LANL2TZ A Los Alamos effective core potential with a triple-zeta valence basis set, with f type polarization functions added on the heavy atoms the potential covers. The f functions describe electron correlation and polarization around heavier elements more accurately than the unaugmented set. (UNVERIFIED) LANL2TZ(f) A correlation-consistent basis set giving 2 functions per valence orbital with a set of polarization functions. It is the smallest member of its family and the cheapest at which correlated methods give meaningful results. It is used as the lowest point of an extrapolation towards the complete basis set limit. (UNVERIFIED) cc-pVDZ A correlation-consistent basis set that gives 3 functions per valence orbital together with polarization functions up to f type. It is substantially more accurate than the double-zeta set for correlated calculations. It is widely used as a balance between accuracy and computational cost. (UNVERIFIED) cc-pVTZ A correlation-consistent basis set that gives 4 functions per valence orbital together with polarization functions up to g type. It is used where a triple-zeta result is not converged enough, and as one point in an extrapolation to the complete basis set limit. Its cost restricts it to modest-sized molecules. (UNVERIFIED) cc-pVQZ A correlation-consistent basis set that gives 5 functions per valence orbital together with polarization functions up to h type. It is built for benchmark calculations that need to come very close to the complete basis set limit for correlation energy. Its cost restricts it to small systems. (UNVERIFIED) cc-pV5Z A correlation-consistent basis set that gives 6 functions per valence orbital together with polarization functions up to i type. It is 1 step larger than the quintuple-zeta set and brings results a further small step closer to the complete basis set limit. In practice it is feasible only for atoms and very small molecules. (UNVERIFIED) cc-pV6Z A correlation-consistent basis set that gives 8 functions per valence orbital together with very extensive polarization functions. It exists to study how results converge towards the complete basis set limit, rather than to be used in ordinary work. Its cost restricts it to single atoms and diatomic molecules. (UNVERIFIED) cc-pV8Z A correlation-consistent basis set that gives 9 functions per valence orbital together with polarization functions of very high angular momentum. It is among the largest sets ever constructed in this family and serves only to extrapolate towards the complete basis set limit on the smallest systems. It is too expensive for nearly all molecular calculations. (UNVERIFIED) cc-pV9Z An auxiliary basis set built to be paired with the correlation-consistent double-zeta orbital basis in resolution-of-identity or density fitting calculations. It approximates products of orbital basis functions rather than representing orbitals itself. Pairing it with a different orbital basis degrades the approximation it was fitted for. (UNVERIFIED) cc-pVDZ-RI An auxiliary basis set built to be paired with the correlation-consistent triple-zeta orbital basis in resolution-of-identity or density fitting calculations. It approximates products of orbital basis functions so integrals can be evaluated faster. It is not used to represent orbitals. (UNVERIFIED) cc-pVTZ-RI An auxiliary basis set built to be paired with the correlation-consistent quadruple-zeta orbital basis in resolution-of-identity or density fitting calculations. It carries the extra functions that fitting products of the larger orbital basis requires. Its purpose is speed, not accuracy. (UNVERIFIED) cc-pVQZ-RI An auxiliary basis set built to be paired with the correlation-consistent quintuple-zeta orbital basis in resolution-of-identity or density fitting calculations. It is not used to represent orbitals; it approximates products of them so that integrals can be evaluated faster. Without it, calculations at that orbital basis size are often impractical. (UNVERIFIED) cc-pV5Z-RI An auxiliary basis set built to be paired with the correlation-consistent sextuple-zeta orbital basis in resolution-of-identity or density fitting calculations. It approximates products of orbital basis functions rather than the orbitals themselves. It is what allows benchmark calculations at that orbital basis size to run at all. (UNVERIFIED) cc-pV6Z-RI A correlation-consistent basis set at double-zeta level that adds tight functions, of large exponent, to describe correlation involving the core electrons. Standard valence sets assume the core is inert, which is a poor approximation for properties that depend on core-valence correlation. It is the smallest member of the family and the cheapest at which the effect can be seen at all. (UNVERIFIED) cc-pCVDZ A correlation-consistent basis set at triple-zeta level that adds tight functions, of large exponent, to describe correlation involving the core electrons. Standard valence sets assume the core is inert, which is a poor approximation for properties that depend on core-valence correlation. It is the level at which core-valence correlation is usually first worth its cost. (UNVERIFIED) cc-pCVTZ A correlation-consistent basis set at quadruple-zeta level that adds tight functions, of large exponent, to describe correlation involving the core electrons. Standard valence sets assume the core is inert, which is a poor approximation for properties that depend on core-valence correlation. It is used for benchmark work where core-valence correlation must be converged rather than merely included. (UNVERIFIED) cc-pCVQZ A correlation-consistent basis set at quintuple-zeta level that adds tight functions describing correlation involving the core electrons, without diffuse functions. It is used for benchmark work where core-valence correlation must be converged. Its size restricts it to very small systems. (UNVERIFIED) cc-pCV5Z A correlation-consistent basis set at sextuple-zeta level that adds tight functions describing correlation involving the core electrons, without diffuse functions. It is among the largest such sets available and approaches the complete basis set limit for the core-valence contribution. Its size restricts it to benchmark work on very small systems. (UNVERIFIED) cc-pCV6Z A correlation-consistent double-zeta basis set carrying 1 extra energy-optimised d type polarization function beyond the standard set. The extra function was added because the standard double-zeta set was found short of d functions for the second-row elements sodium to argon. For first-row elements it behaves like the ordinary double-zeta set. (UNVERIFIED) cc-pV(D+d)Z A correlation-consistent triple-zeta basis set carrying 1 extra optimised d type polarization function beyond the standard set. It describes polarization and correlation for the second-row elements sodium to argon more completely than the ordinary triple-zeta set. For other elements the 2 are equivalent. (UNVERIFIED) cc-pV(T+d)Z A correlation-consistent quadruple-zeta basis set carrying 1 extra optimised d type polarization function beyond the standard set. It is used for benchmark-quality results on the second-row elements sodium to argon, where d function flexibility limits accuracy. For other elements the 2 are equivalent. (UNVERIFIED) cc-pV(Q+d)Z A correlation-consistent quintuple-zeta basis set carrying 1 extra optimised d type polarization function beyond the standard set. It is used in high-level benchmark work on small molecules containing the second-row elements sodium to argon. For other elements the 2 are equivalent. (UNVERIFIED) cc-pV(5+d)Z A correlation-consistent sextuple-zeta basis set carrying 1 extra optimised d type polarization function beyond the standard set. It is a highly specialised set for work on the second-row elements sodium to argon near the complete basis set limit. For other elements the 2 are equivalent. (UNVERIFIED) cc-pV(6+d)Z A correlation-consistent basis set giving 2 functions per valence orbital, with 1 diffuse function added at each angular momentum present. The diffuse functions matter for anions, weakly bound complexes and response properties. It is the smallest augmented member of the correlation-consistent family. (UNVERIFIED) aug-cc-pVDZ A correlation-consistent basis set giving 3 functions per valence orbital, with a set of diffuse functions added. The diffuse functions improve the description of electron density far from the nuclei, which matters for anions, weak interactions and polarizabilities. It is a widely used standard for accurate calculations that need diffuse functions. (UNVERIFIED) aug-cc-pVTZ A correlation-consistent basis set giving 4 functions per valence orbital with a set of diffuse functions added at each angular momentum. The diffuse functions matter for anions, weakly bound complexes and response properties. It is a common point in an extrapolation towards the complete basis set limit. (UNVERIFIED) aug-cc-pVQZ A correlation-consistent basis set that gives 5 functions per valence orbital and adds a set of diffuse functions. The diffuse functions matter when electron density reaches far from the nuclei, as in an anion or a weakly bound complex. It is a benchmark-level set and its cost limits it to very small systems. (UNVERIFIED) aug-cc-pV5Z A correlation-consistent basis set that gives 6 functions per valence orbital and adds a set of diffuse functions. It is one of the largest standard augmented sets available. It is used to study convergence, or for properties that need extreme accuracy in both correlation and diffuse electron density. (UNVERIFIED) aug-cc-pV6Z A correlation-consistent basis set that gives 8 functions per valence orbital and adds a set of diffuse functions. It is a research set used almost only on atoms and diatomic molecules, to study how properties sensitive to diffuse electrons converge towards the complete basis set limit. It is too expensive for calculations on larger molecules. (UNVERIFIED) aug-cc-pV8Z A correlation-consistent basis set that gives 9 functions per valence orbital and adds a set of diffuse functions. Like the octuple-zeta version, it is a research set used almost only on atoms and diatomic molecules, to study convergence towards the complete basis set limit. It is too expensive for calculations on larger molecules. (UNVERIFIED) aug-cc-pV9Z An auxiliary basis set built to be paired with the diffuse-augmented correlation-consistent double-zeta orbital basis in resolution-of-identity or density fitting calculations. It carries the extra functions needed to fit products involving the diffuse part of that basis. Pairing it with a different orbital basis degrades the approximation it was fitted for. (UNVERIFIED) aug-cc-pVDZ-RI An auxiliary basis set built to be paired with the diffuse-augmented correlation-consistent triple-zeta orbital basis in resolution-of-identity or density fitting calculations. It carries the extra functions needed to fit products involving the diffuse part of that basis. Pairing it with a different orbital basis degrades the approximation it was fitted for. (UNVERIFIED) aug-cc-pVTZ-RI An auxiliary basis set built to be paired with the diffuse-augmented correlation-consistent quadruple-zeta orbital basis in density fitting calculations. It carries the extra functions needed to fit products involving the diffuse part of that basis. Its purpose is speed, not accuracy. (UNVERIFIED) aug-cc-pVQZ-RI An auxiliary basis set built to be paired with the diffuse-augmented correlation-consistent quintuple-zeta orbital basis in resolution-of-identity or density fitting calculations. It carries the extra functions needed to fit products involving the diffuse part of that orbital basis. Its purpose is speed, not accuracy. (UNVERIFIED) aug-cc-pV5Z-RI An auxiliary basis set built to be paired with the diffuse-augmented correlation-consistent sextuple-zeta orbital basis in resolution-of-identity or density fitting calculations. It is required in practice, because calculations at that orbital basis size are otherwise not feasible. Its purpose is speed, not accuracy. (UNVERIFIED) aug-cc-pV6Z-RI A correlation-consistent basis set at double-zeta level that adds both diffuse functions and tight functions describing correlation involving the core electrons. Standard valence sets assume the core is inert, which fails for properties that depend on it. It is the smallest combination that carries both additions. (UNVERIFIED) aug-cc-pCVDZ A correlation-consistent basis set at triple-zeta level that adds both diffuse functions and tight functions describing correlation involving the core electrons. Standard valence sets assume the core is inert, which fails for properties that depend on it. It is the usual working choice where a system needs diffuse character and core-valence correlation at once. (UNVERIFIED) aug-cc-pCVTZ A correlation-consistent basis set at quadruple-zeta level that adds both diffuse functions and tight functions describing correlation involving the core electrons. Standard valence sets assume the core is inert, which fails for properties that depend on it. It is used where both effects must be converged rather than merely present. (UNVERIFIED) aug-cc-pCVQZ A correlation-consistent basis set at quintuple-zeta level that adds both diffuse functions and tight functions describing correlation involving the core electrons. Standard valence sets assume the core is inert, which fails for properties that depend on it. It is a benchmark set whose cost restricts it to small systems. (UNVERIFIED) aug-cc-pCV5Z A correlation-consistent basis set at sextuple-zeta level that adds both diffuse functions and tight functions describing correlation involving the core electrons. Standard valence sets assume the core is inert, which fails for properties that depend on it. It is among the largest standard basis sets available and is used only for benchmark work on very small systems. (UNVERIFIED) aug-cc-pCV6Z A correlation-consistent double-zeta basis set that adds both a set of diffuse functions and 1 extra optimised d type polarization function. It is meant for the second-row elements sodium to argon in situations that also need diffuse functions, such as anions and weakly bound complexes. It combines 2 separate corrections to the standard double-zeta set. (UNVERIFIED) aug-cc-pV(D+d)Z A correlation-consistent triple-zeta basis set that adds both a set of diffuse functions and 1 extra optimised d type polarization function. It targets the second-row elements sodium to argon where both diffuse electron density and d orbital flexibility matter, such as anions and intermolecular complexes. It is the triple-zeta member of that combined family. (UNVERIFIED) aug-cc-pV(T+d)Z A correlation-consistent quadruple-zeta basis set that adds both a set of diffuse functions and 1 extra optimised d type polarization function. It is used for benchmark results on systems containing the second-row elements sodium to argon whose properties depend on both corrections. It is large enough that its use is restricted to small systems. (UNVERIFIED) aug-cc-pV(Q+d)Z A correlation-consistent quintuple-zeta basis set that adds both a set of diffuse functions and 1 extra optimised d type polarization function. It is a specialised benchmark set for the second-row elements sodium to argon in processes that also need diffuse functions. It is close to the complete basis set limit for those elements. (UNVERIFIED) aug-cc-pV(5+d)Z A correlation-consistent sextuple-zeta basis set that adds both a set of diffuse functions and 1 extra optimised d type polarization function, the latter for the second-row elements sodium to argon. It is among the largest combinations available and is used only for extreme benchmark work on very small systems containing those elements. (UNVERIFIED) aug-cc-pV(6+d)Z A correlation-consistent double-zeta basis set that adds diffuse functions together with tight functions weighted towards core-valence correlation. The weighting concentrates the added functions where core and valence electrons actually correlate, rather than describing the core itself. It suits a small system needing both diffuse character and core-valence correlation. (UNVERIFIED) aug-cc-pwCVDZ A correlation-consistent triple-zeta basis set that adds diffuse functions together with weighted core-valence correlating functions. It is the level at which core-valence effects are usually first worth their cost. Both additions are needed for properties sensitive to diffuse electrons in systems where the core cannot be treated as inert. (UNVERIFIED) aug-cc-pwCVTZ A correlation-consistent quadruple-zeta basis set that adds diffuse functions together with weighted core-valence correlating functions. It is used for high-accuracy work where both effects must be converged rather than merely present. Its size restricts it to small molecules. (UNVERIFIED) aug-cc-pwCVQZ A correlation-consistent quintuple-zeta basis set that adds diffuse functions together with weighted core-valence correlating functions. It is a benchmark set used to approach the complete basis set limit. Its cost restricts it to atoms and very small molecules. (UNVERIFIED) aug-cc-pwCV5Z A polarization-consistent basis set at the lowest level of the pc-n family, with diffuse functions added. The family is built so results improve smoothly as the level rises, which makes extrapolation to the limit reliable. This level is the starting point of that sequence rather than a basis for production work. (UNVERIFIED) aug-pc-0 A polarization-consistent basis set at the first level of the pc-n family, with diffuse functions added. The level corresponds roughly to polarized double-zeta quality. It is the smallest member of the family used for real calculations. (UNVERIFIED) aug-pc-1 A polarization-consistent basis set at the second level of the pc-n family, with a set of diffuse functions added. The second level corresponds to polarized triple-zeta quality, and the family is built so that results converge smoothly as the level rises. The diffuse functions make it suitable for anions, intermolecular interactions and polarizabilities. (UNVERIFIED) aug-pc-2 A polarization-consistent basis set at the third level of the pc-n family, with a set of diffuse functions added. The third level corresponds to polarized quadruple-zeta quality. It is used where both high accuracy and a correct description of diffuse electron density are needed. (UNVERIFIED) aug-pc-3 A polarization-consistent basis set at the fourth level of the pc-n family, with a set of diffuse functions added. The fourth level corresponds to polarized quintuple-zeta quality and is the largest in common use in this family. It is a benchmark set for small systems whose properties depend on diffuse electrons. (UNVERIFIED) aug-pc-4 An auxiliary basis set developed by the Ahlrichs group to fit the Coulomb part of the electron repulsion in a density fitting calculation. Fitting only the Coulomb term, rather than the whole 2-electron integral set, gives most of the speed-up at very little cost in accuracy. It is matched to the Karlsruhe orbital basis sets. (UNVERIFIED) Ahlrichs Coulomb fitting A set of polarization functions, of higher angular momentum than the atom occupies in its ground state, developed by the Ahlrichs group for the Karlsruhe basis sets. Polarization functions let the electron density distort away from atomic symmetry, which is what bonding requires. Without them a basis set cannot describe a bond angle correctly. (UNVERIFIED) Ahlrichs Polarization A basis set from the Ahlrichs group giving 2 functions per valence orbital together with polarization functions. The polarization functions let the electron cloud distort away from atomic symmetry, which is what bonding requires and which the unpolarized member of the family cannot do. It is the smallest basis set in that family suitable for real calculations. (UNVERIFIED) Ahlrichs pvDZ A basis set from the Ahlrichs group giving 2 functions per valence orbital and no polarization functions. Without polarization functions the electron cloud cannot distort away from atomic symmetry, so bond angles are poorly described. It is used as a starting point rather than for results. (UNVERIFIED) Ahlrichs VDZ A basis set from the Ahlrichs group giving 3 functions per valence orbital and no polarization functions. The extra valence function improves the radial description while leaving the angular one unchanged. Like its double-zeta counterpart it is normally used with polarization functions added. (UNVERIFIED) Ahlrichs VTZ A type of quantum mechanics basis set consisting of periodic sine and cosine functions (waves) defined by their direction and wavelength. Plane waves are particularly well-suited for calculations on periodic systems like crystals or surfaces, often used in conjunction with Density Functional Theory and pseudopotentials. Their systematic improvability by increasing the number of waves (controlled by an energy cutoff) is a key advantage in solid-state simulations. (UNVERIFIED) plane-wave A method that represents electronic wavefunctions or densities as values on a grid of points in ordinary 3-dimensional space, rather than as coefficients of basis functions. Accuracy is then improved by making the grid finer, which is a single adjustable control. It parallelises well, because each point depends mostly on its neighbours. (UNVERIFIED) real-space A coordinate analysis that follows the steepest-descent path, in mass-weighted coordinates, downhill from a transition state to the reactant and product minima on the potential energy surface. The resulting path is the minimum energy path that connects reactants and products through that transition state. It shows which minima a transition state connects and describes the reaction at 0 kelvin, without thermal motion. (UNVERIFIED) intrinsic reaction coordinate analysis An analysis method that locates the highest point along the lowest-energy path joining reactant and product. That point is the transition state, and its height above the reactant is the reaction barrier. Finding it is what turns a computed path into a predicted rate. (UNVERIFIED) intrinsic reaction coordinate maximum analysis A form of reaction analysis focused on characterizing the features of the potential energy surface (PES), which maps the system's energy as a function of its atomic coordinates. This analysis identifies critical points like energy minima (stable states), saddle points (transition states), and intersections between surfaces that govern molecular stability and reactivity. It helps visualize the energetic landscape explored during reactions or conformational changes. surface analysis A surface analysis that locates and characterises conical intersections, the molecular geometries at which 2 electronic potential energy surfaces of the same spin multiplicity have the same energy. At these points a molecule can pass very quickly from 1 electronic state to another without emitting light, which is central to photochemical reactions and to relaxation after light absorption. Their location and shape determine how efficient such transitions are. (UNVERIFIED) conical intersection point analysis A surface analysis that locates saddle points on a potential energy surface: stationary points, where all forces are zero, at which the energy is a maximum along 1 direction, the reaction path, and a minimum along all other directions. Such a point is the highest point on the lowest-energy path between reactant and product. Locating it identifies the transition state and gives the activation energy of the reaction. (UNVERIFIED) saddle analysis A surface analysis that locates the minimum energy crossing point, the lowest-energy molecular geometry at which 2 potential energy surfaces of different spin states, such as a singlet and a triplet, have the same energy. Such points control the pathways of spin-forbidden reactions and of intersystem crossing, the change of a molecule from 1 spin state to another. Locating them helps to explain how a molecule changes spin state. (UNVERIFIED) minimum crossing analysis A form of reaction analysis aimed at identifying and characterizing the sequence of intermediate states and transition states that connect reactants to products, effectively mapping the reaction pathway. This can involve exploring multiple possible routes, using algorithms to find minimum energy paths, or analyzing trajectories from reaction dynamics simulations. pathway analysis A pathway analysis that follows gradient extremals on a potential energy surface, which are curves along which the energy gradient points along an eigenvector of the matrix of second derivatives (the Hessian). Such curves pass through minima and saddle points, so following one uphill from a minimum can lead to a transition state. (UNVERIFIED) trace gradient extremal analysis A category for file formats designed to store numerical data in a specific structure that is easily parsed and read by graphing or plotting software (e.g., Grace, Gnuplot). These are structured data files typically containing columns of numbers and plotting directives. plot data format A text-based data file format for the Grace/Xmgrace plotting program. It contains numerical data along with commands that specify the appearance of the graph, such as titles, labels, and colors. This format allows for the creation of complex, publication-quality plots. (UNVERIFIED) agr format A file format that represents a molecular system's topology after certain atoms, typically solvent or ions, have been removed. This format is created to simplify analysis by focusing only on the solute, while retaining the necessary parameter information for calculations. It is commonly used in post-processing steps like MM/PBSA free energy calculations. (UNVERIFIED) stripped topology format A text data format in which cpptraj writes time-series results by default, as columns of numbers with the frame or time in the first column and the calculated values in the rest. It is space-delimited and carries no header structure beyond a comment line, so plotting and analysis programs read it directly. Files are commonly given the extension .dat. (UNVERIFIED) Typically with .dat extension, but could be using other file extension. CPPTRAJ standard output format A data file format designed to be used with the Gnuplot plotting program. It typically consists of columns of numerical data in a simple text-based layout that Gnuplot can easily parse and visualize. This format is widely used for generating scientific graphs and plots from raw data. (UNVERIFIED) gnu format Gnuplot data format A file format associated with the XPLOR-NIH software, used for representing 3D grid data. It is a text-based format that defines a grid and the data values at each point, often used for electron density maps. This format is a common way to store volumetric data in crystallography and NMR. (UNVERIFIED) XPLOR format A file format for representing data that is distributed over a three-dimensional grid. Originating from the XPLOR-NIH software, it is used to store volumetric data like electron density or potential maps. This format is used in structural biology to visualize grid-based properties around a molecule. .grid grid format A file format for the OpenDX (Data Explorer) visualization program, used to represent 3D grid-based data. It stores scalar or vector data on a regular grid, making it suitable for visualizing properties like electrostatic potential or molecular density. This format is a common output for scientific programs that generate volumetric data. (UNVERIFIED) OpenDX format DX format A data format that stores the velocity of every atom of a system at each saved step of an AMBER simulation. It records velocities where a trajectory format records positions. It is needed for analyses such as velocity autocorrelation that cannot be done from coordinates alone. (UNVERIFIED) AMBER mdvel format mdvel format A text data format in which AMBER records each exchange attempt of a replica exchange simulation, listing for every replica values such as its temperature, its potential energy, its condition before and after the attempt and the running success rate. The file is named rem.log by default. It is read to check how often exchanges succeed and how well the replicas mix. (UNVERIFIED) Amber Replica Exchange Log rem.log AMBER REM log format A file format that holds the eigenvectors and eigenvalues produced by principal component analysis, normal mode analysis or quasiharmonic analysis in AMBER. The vectors come from diagonalising a covariance or mass-weighted covariance matrix, and are used in essential dynamics analysis of molecular simulations. There is no required extension: the file is usually called evecs.dat or evecs.out, and its layout is standardised so that AMBER tools such as cpptraj and ptraj can read and write it. (UNVERIFIED) evecs.dat evecs.out AMBER eigenvector format A text data format in which AMBER records the protonation state of each titratable residue at each Monte Carlo step of a constant pH simulation, with full records that also give the solvent pH, the time step and the time. The protonated fractions and pKa values of the residues are later calculated from it, for example with the cphstats program. It is an output of the simulation and is not read to start or restart one. (UNVERIFIED) .cpout AMBER cpout format cpout format A computational procedure within structural analysis used to find the atomic coordinates corresponding to a stationary point, typically a local minimum, on the potential energy surface, representing a stable or equilibrium structure. Starting from an initial guess geometry, algorithms iteratively adjust atomic positions to minimize the energy and reduce the forces acting on the atoms until convergence criteria are met. The result is a stable structure near the starting geometry, which need not be the most stable structure overall. (UNVERIFIED) geometry optimization analysis A specific geometry optimization technique designed for large molecular systems that employs the Fragment Molecular Orbital (FMO) method. The FMO method divides the large system (like a protein or DNA segment) into smaller, manageable fragments, performs quantum mechanical calculations on fragments and fragment pairs, and combines the results to approximate the properties (energy, gradient) of the entire system. This fragmentation allows for efficient geometry optimization of very large biomolecules at a quantum mechanical level. (UNVERIFIED) FMO geometry optimization analysis An energy analysis calculation that determines the potential energy of a molecular system for a single, fixed geometric arrangement of its atoms (a specific 'snapshot' or structure), without performing geometry optimization or simulating dynamics. This provides the energy value for that particular conformation according to the chosen theoretical method (e.g., force field or quantum mechanical level). It's often used to compare the relative energies of different pre-defined structures. single point energy analysis An energy analysis that computes potential energies with a high-level and computationally expensive treatment, such as coupled cluster theory or density functional theory with a large basis set. It is applied to a few chosen structures, typically optimised minima, transition states or representative conformations, rather than to a whole trajectory. The values serve as benchmarks for reaction barriers, binding energies and conformational stabilities obtained more cheaply. (UNVERIFIED) high accuracy energies analysis An energy analysis technique designed to partition the total interaction energy between two or more interacting molecular fragments (e.g., in a complex) into physically meaningful components based on the underlying quantum mechanical or molecular mechanical framework. Common components include electrostatics, exchange-repulsion (Pauli repulsion), polarization (induction), and charge transfer (or dispersion in some schemes), providing insights into the nature of the forces driving the interaction. Different EDA schemes exist (e.g., Morokuma, SAPT, FMO-EDA). (UNVERIFIED) energy decomposition analysis A structural analysis technique that involves calculating the forces acting on each atom for a given molecular configuration. These forces represent the negative gradient of the potential energy and indicate the direction and magnitude of the tendency for atoms to move to lower the system's energy. Analyzing forces is crucial for geometry optimization (where forces are minimized), understanding structural strain, and driving molecular dynamics simulations. (UNVERIFIED) force analysis A binary data format that stores 3-dimensional volumetric data on a regular grid, most often electron density from X-ray crystallography or a map from cryogenic electron microscopy. It is closely related to, and largely compatible with, the MRC format. It lets the shape of a protein or nucleic acid be seen and interpreted. (UNVERIFIED) .ccp4 CCP4 map format A binary data format in which cpptraj stores the pre-calculated pairwise distances between trajectory frames, so that a cluster analysis can reuse them instead of computing them again. It has no mandatory extension, but .cmatrix is the conventional one. (UNVERIFIED) .cmatrix CPPTRAJ pairwise distance cache format cmatrix format A type of analysis, often performed using data from simulations or specialized calculations complementary to MD, focused on determining the thermodynamic properties governing molecular systems. This includes calculating state functions like enthalpy, entropy, and particularly free energy, which dictate stability, equilibrium constants, and the spontaneity of molecular processes. Such analysis provides insights into the energetic driving forces behind biomolecular behavior. (UNVERIFIED) thermodynamics analysis A thermodynamics analysis that deals with how a system's properties change with temperature or involve heat exchange. This includes calculating quantities like heat capacity (how much energy is needed to raise the temperature), coefficients of thermal expansion, or enthalpy changes associated with specific processes. These properties are important for understanding temperature dependence and energy balance in molecular systems. (UNVERIFIED) thermal property analysis A specific type of thermal properties analysis focused on quantifying the heat absorbed or released during chemical reactions or phase transitions, often involving the calculation of enthalpies of formation, reaction enthalpies, or activation enthalpies. Typically employing quantum mechanical methods or specialized simulation setups, this analysis provides fundamental data on the energetics of chemical transformations. It helps predict reaction feasibility and energy output/input. (UNVERIFIED) thermochemical analysis A text data format used by AMBER to define molecular building blocks, giving for each its atoms, their types and charges, the bonds between them and its geometry, as either internal or Cartesian coordinates. It is a common way to introduce a residue that no standard library contains, and 1 file can hold many building blocks. Such files are usually written by tools such as antechamber rather than by hand. (UNVERIFIED) .prep .prepi AMBER prep format prep format A text data format used by AMBER to store a library of molecular building blocks, giving for each its atoms, their types and charges, its bonds, its coordinates and the atoms through which it joins its neighbours. A single file can hold many building blocks, and also parameter sets. The standard residue libraries of AMBER are distributed in it, usually with the extension .lib or .off. (UNVERIFIED) .off AMBER Object File Format AMBER off format AMBER lib format OFF format A thermodynamics analysis method that computes the free energy of a system, or more often the free energy difference between 2 states such as bound and unbound or folded and unfolded. Free energy combines energy and disorder, so it, rather than energy alone, decides which state a process settles into. Obtaining it usually needs enhanced sampling, because the states of interest are separated by barriers. (UNVERIFIED) free energy calculation analysis A specific computational task within a free energy calculation framework that uses the Fragment Molecular Orbital (FMO) method to determine the energy and properties of a molecule in its unbound ('free') state. This is often required when calculating binding free energies, where the energy of the isolated components serves as a reference point (G free). Employing FMO makes such quantum mechanical calculations feasible for the large molecules often involved in biomolecular binding events. (UNVERIFIED) free state FMO analysis A free energy analysis method that computes the solvation free energy from a 3D reference interaction site model (3D-RISM) calculation using the Gaussian fluctuation expression. This expression is not tied to a particular closure, the approximate relation used to solve the integral equations, and is used in place of the closure's own formula. It has been shown to improve absolute solvation free energies for both polar and non-polar solutes, but not necessarily relative ones. (UNVERIFIED) Gaussian fluctuation free energy analysis A free energy analysis method that corrects the solvation free energy computed by the 3D reference interaction site model (3D-RISM), which tends to overestimate its non-polar part, with a pressure correction based on the partial molar volume of the solute. Unlike the universal correction, it has no fitted parameters. (UNVERIFIED) PC+ correction analysis A free energy analysis method that corrects the solvation free energy computed by the 3D reference interaction site model (3D-RISM) with a term based on the partial molar volume of the solute, using parameters fitted for the particular solvent model. It counters the tendency of 3D-RISM to overestimate the non-polar part of the solvation free energy, and has been used almost only for water at ambient conditions. (UNVERIFIED) UC Correction universal correction analysis A semi-empirical quantum method derived from density functional theory by expanding its energy about a reference density and tabulating the resulting integrals in advance. Tabulating them removes almost all the cost while keeping a quantum description. It is 2 to 3 orders of magnitude faster than the theory it comes from. (UNVERIFIED) DFTB density-functional based tight-binding A density-functional based tight-binding method that solves for the atomic charges self-consistently, so each atom's charge responds to the others rather than being fixed. Adding that step corrects the description of polar and charged systems, which the non-self-consistent form handles badly. (UNVERIFIED) SCC-DFTB self-consistent-charge density-functional based tight-binding A tight binding method that carries the expansion of the energy one order further than the self-consistent charge form, adding a term for how an atom's response to charge itself changes with charge. That term corrects the description of charged and hydrogen-bonded systems, which the lower order handles poorly. It is the form recommended for biomolecular work. (UNVERIFIED) DFTB3 third-order self-consistent-charge density-functional based tight-binding An interaction that cannot be described correctly without quantum mechanics, because it depends on the wave nature of electrons or on energy taking discrete values. Bond formation, charge transfer and dispersion are all of this kind. A classical force field represents their effects approximately with fitted terms rather than describing them. (UNVERIFIED) quantum interaction A quantum mechanical interaction representing the coupling between the magnetic moment associated with the spin of an electron and the magnetic moment of a nearby atomic nucleus. This interaction leads to small shifts and splittings ('hyperfine structure') in atomic and molecular energy levels, which are observable in high-resolution spectroscopy, particularly Electron Paramagnetic Resonance (EPR) for species with unpaired electrons, and also in NMR. It provides valuable information about the electron distribution (spin density) near the nucleus and thus molecular structure. (UNVERIFIED) hyperfine coupling A relativistic quantum mechanical interaction between a particle's intrinsic spin angular momentum and its orbital angular momentum (due to its motion, e.g., an electron orbiting a nucleus). This coupling mixes electronic states and becomes increasingly significant for electrons in atoms with high atomic number (heavy atoms). Spin-orbit coupling influences electronic energy levels, causes splitting of spectral lines (fine structure), and facilitates transitions between states of different spin multiplicity (e.g., intersystem crossing in photochemistry, important for phosphorescence). (UNVERIFIED) spin-orbit coupling A physical effect arising from the intrinsic angular momentum carried by electrons and atomic nuclei. Spin gives each particle a magnetic moment, and the interactions of these moments with magnetic fields, with each other and with orbital motion produce magnetic resonance signals and the fine and hyperfine structure of spectra. A calculation that ignores spin cannot reproduce them. (UNVERIFIED) spin effect A quantum mechanical method that obtains the energy and other properties of a system from its electron density rather than from the many-electron wavefunction. The density depends on 3 spatial coordinates however many electrons there are, which is what makes the method affordable for large systems. Its accuracy rests on the approximate functional chosen to relate density to energy. (UNVERIFIED) DFT density functional theory A QM/MM scheme in which the partial charges of the classical region enter the quantum calculation directly, so the quantum region's electron distribution responds to its surroundings. That response is what captures how a protein environment alters the chemistry inside an active site. It costs more than the mechanical scheme and is markedly more accurate. (UNVERIFIED) electrostatic embedding QM/MM An Ewald summation method for a periodic system in which part is described quantum mechanically and the rest by a force field. It must handle 3 kinds of interaction: within the quantum region, within the classical region, and between them. Getting the third right is what distinguishes it, because the 2 descriptions represent charge differently. (UNVERIFIED) Ewald for QM/MM QM/MM Ewald summation A QM/MM method using more than 2 levels of description, such as an expensive quantum method for a reacting group, a cheaper quantum method around it, and a force field beyond that. Each boundary must be handled separately, so the arrangement is more delicate than a 2-level one. It is used where the reacting region is small but its immediate surroundings still need quantum treatment. (UNVERIFIED) multilayer QM/MM A QM/MM method in which solvent molecules change gradually between quantum and classical treatment as they cross a transition shell around the quantum region, with their partial classical character set by their distance from its centre. The energy is a weighted average over the ways of assigning the molecules in the shell to the quantum or the classical region, with weights that change smoothly as the molecules move. This avoids the sudden jumps in force that would occur if a molecule switched abruptly between the 2 descriptions. (UNVERIFIED) difference-based adaptive solvation QM/MM A QM/MM method in which the quantum region is redefined by distance as the simulation runs, and the forces on its atoms come from a larger quantum calculation that also includes a surrounding buffer region. The buffer keeps the forces on the quantum atoms from being distorted by the nearby boundary, while the atoms outside take their forces from a second calculation with a smaller quantum region. Because the forces do not come from a single energy function, energy is not conserved and a thermostat is used to hold the temperature. (UNVERIFIED) adaptive buffered force-mixing A structural analysis method that computes a field, most often the electrostatic potential, and displays it on the surface of a molecule. Colouring the surface by potential shows at a glance which regions would attract a positive or a negative partner. It is how electrostatic complementarity between 2 binding partners is judged. (UNVERIFIED) surface potential analysis An analysis method that computes the electrostatic potential a molecule's charge distribution produces at points around it. Mapped onto the molecular surface, it shows where an approaching charge would be attracted or repelled. It is how a binding site is judged to be electrostatically complementary to a ligand. (UNVERIFIED) electrostatic potential calculation analysis An analysis method that calculates the vibrational frequencies and normal modes (patterns of atomic motion) of a molecule at a stationary point on its potential energy surface, such as an energy minimum or a transition state. The frequencies can be compared with measured vibrational spectra, such as infrared or Raman spectra, to check a structure or a force field. They also show what kind of stationary point was found: a minimum has only real frequencies, and a transition state has exactly 1 imaginary frequency. (UNVERIFIED) frequency analysis A refinement within frequency analysis that goes beyond the standard harmonic approximation (which assumes vibrations occur in a perfectly parabolic potential well) to account for the true, anharmonic shape of the potential energy surface. Calculating these corrections yields more accurate vibrational frequencies that better match experimental observations, particularly for large-amplitude motions or systems with significant vibrational coupling. This improves the quantitative interpretation of vibrational spectra and thermodynamic properties derived from frequencies. (UNVERIFIED) anharmonic vibrational corrections analysis A simulation method in which the forces at each step come from a fast approximate quantum mechanical calculation, solved fully at each new geometry. Being quantum, it can describe bonds being made and broken, which a force field cannot. Being approximate, it is fast enough to run for a useful length of time. (UNVERIFIED) SEBOMD semi-empirical Born-Oppenheimer molecular dynamics A simulation method that treats the region where bonds form and break with the ReaxFF reactive force field and the rest of the system with a conventional AMBER force field. The reactive force field costs far more, so confining it to a small region makes a large system affordable. It works in a similar way to QM/MM, with a reactive force field in place of quantum mechanics. (UNVERIFIED) hybrid ReaxFF/AMBER molecular dynamics A force field that represents how a molecule's charge distribution is distorted by the electric field around it, rather than fixing the charges in advance. It matters where the environment changes markedly, as when a ligand moves from water into a protein interior. It costs several times more, because the induced response must be solved for at every step. (UNVERIFIED) fluctuating charge model polarizable force field A polarizable force field in which each atom carries a multipole expansion built from Gaussian-shaped charge distributions, and responds to the field around it. Using distributed shapes rather than points lets the model describe how charge density overlaps between close atoms. It targets the short-range region where point-charge models are least reliable. (UNVERIFIED) polarizable Gaussian multipole force field A correction added to a quantum chemistry calculation to supply the weak long-range attraction that most density functionals omit. The functionals fail here because the attraction arises from correlated fluctuations they do not describe. The correction is a fitted term depending on the atoms and their separation, and it is what makes density functional theory usable for stacking and binding. (UNVERIFIED) D3 dispersion model grimme dispersion correction empirical dispersion correction A Gaussian-shaped constraint simulation is a type of restrained simulation where the artificial potential used to guide the system has the mathematical form of a Gaussian function. This provides a smooth, harmonic-like restraint near the target value but becomes weaker further away. Gaussian-shaped constraint simulation A force field parameterization method that obtains the Lennard-Jones parameters between 2 different atom types from those of each type alone, taking the arithmetic mean of their sizes and the geometric mean of their well depths. Without such a rule every pair of atom types would need its own fitted parameters. The rule is a convenience rather than physics, which is why some pairs need an explicit override. (UNVERIFIED) Lorentz-Berthelot mixing modeling A protocol for deriving atomic charges in which the electrostatic potential is computed at the Hartree-Fock level with the 6-31G* basis set, then point charges are fitted to it with a restraint pulling buried charges towards zero. That combination is not the most accurate available; it is used because it systematically overestimates polarity by about the amount a simulation in fixed-charge water requires. It is the convention behind many AMBER force fields, including the general AMBER force fields, so charges for a new molecule used with them are usually derived the same way. (UNVERIFIED) HF/6-31G RESP A coarse-grained force field for proteins and nucleic acids whose beads are chosen to preserve the chemical character and the directionality of interactions such as hydrogen bonds. It keeps more chemical detail than a purely shape-based mapping, at a higher cost. It can be combined with explicit coarse-grained water. (UNVERIFIED) SIRAH force field A force field parameterization method that derives parameters for a metal ion and its coordination site by fitting them to reproduce experimental data for a set of known metal complexes, such as crystal structures, vibrational frequencies or binding energies. Metal ions often need this treatment because general-purpose parameterization methods do not capture their electronic structure and bonding well. (UNVERIFIED) empirical metal center parameterization modeling A free energy analysis method that estimates binding free energy using explicit water molecules near the solute and a continuum model for the bulk. It keeps the specific water-mediated contacts that a continuum model cannot represent, while avoiding the cost of explicit solvent everywhere. The energies are then combined in the usual molecular mechanics Poisson-Boltzmann surface area way. (UNVERIFIED) hybrid solvent MM-PBSA analysis A free energy analysis method that recomputes the electrostatic (decharging) step of an alchemical binding free energy calculation with a Poisson-Boltzmann continuum model, combining the energies of the snapshots with the Bennett acceptance ratio. Raising the solute dielectric constant in the continuum model lets it include the screening by electronic polarization that a standard non-polarizable force field leaves out. The atomic radii are first calibrated so that the continuum model reproduces the explicit-solvent results. (UNVERIFIED) BAR/PBSA analysis Kinetic map analysis is a method used to visualize the network of transitions between different conformational states that a molecule visits during a simulation. It represents the molecular states as nodes and the transitions between them as connections, often colored or weighted by their probability or rate. This provides a clear and intuitive map of the dominant pathways for conformational change, helping to understand the mechanism and timescales of processes like ligand binding or protein folding. kinetic map analysis Commute map analysis is a method used to identify how different parts of a molecule are dynamically connected, even if they are far apart. It transforms complex molecular coordinates into a mathematical space where the Euclidean distance between points directly corresponds to the commute time (the mean time it takes for a molecule to transition between two conformations and back). This analysis helps to uncover long-range communication networks within a protein that are essential for its function, such as in allosteric regulation. commute map analysis A thermodynamics analysis method that splits the total entropy change of a process into separate contributions. The split may be by kind of motion, into translational, rotational and vibrational parts, or by component, into solute and solvent. It shows which source of disorder drives a process, for example whether binding is favoured by the release of ordered water. (UNVERIFIED) entropic decomposition analysis An advanced analysis method that attempts to generate a 3-dimensional model of a molecule or a molecular complex using statistical information about its structure. It takes one or more distribution functions, such as the pairwise distances between atoms obtained from a simulation or experiment, and uses an optimization algorithm to find a 3D structure that is consistent with those distributions. This technique can be used to build structural models from sparse or low-resolution data. (UNVERIFIED) molecular reconstruction from distribution functions analysis A structural analysis method that separates the change in shape of a group of atoms between 2 conformations into a pure rotation and an internal deformation. It does this by factorising the matrix that maps one set of coordinates onto the other. It answers whether a domain has turned as a rigid body or genuinely changed shape. (UNVERIFIED) polar decomposition analysis A simulation method that includes quantum behaviour of the atomic nuclei themselves, by replacing each nucleus with a ring of connected copies. The spread of the ring represents the uncertainty in the particle's position, which is large for light atoms. It matters for hydrogen, where treating the nucleus as a point misplaces bond lengths and reaction barriers. (UNVERIFIED) path integral molecular dynamics Replica-exchange free energy perturbation is an enhanced sampling method that combines replica-exchange with an alchemical free energy calculation. It runs multiple simulations in parallel, each at a different intermediate state (lambda value) of the alchemical transformation, and periodically attempts to swap the lambda values between replicas. replica-exchange free energy perturbation A simulation method in which groups can gain or lose an electron during the run, coupled to a chosen electrode potential. It is the redox counterpart of constant pH simulation and uses the same machinery. It is used for proteins whose function involves moving electrons, where the oxidation state is not fixed in advance. (UNVERIFIED) constant redox potential molecular dynamics X-ray refinement is the computational process of building and optimizing an atomic model of a molecule to fit the experimental data from an X-ray crystallography experiment. It is an iterative procedure where the atomic coordinates are adjusted to best match the calculated electron density map, while simultaneously ensuring that the model has realistic chemical geometry, such as proper bond lengths and angles. This process is the final and crucial step in determining a high-resolution crystal structure. X-ray refinement A structural analysis method that divides the simulation box into a regular 3-dimensional grid of cells and counts how often atoms of a chosen type fall in each cell across a trajectory. The counts give the average spatial density of those atoms. It is the usual way to map where water or ions gather around a solute. (UNVERIFIED) grid binning analysis A fundamental trajectory post-processing step used to correct the visual representation of molecules in a periodic system. When a molecule crosses a periodic boundary, it can appear 'broken' or jump across the box in the raw output. This procedure translates the atomic coordinates back into the primary simulation box, implicitly fixing broken 'image bonds' to ensure target molecules remain whole and centered. This is a mandatory prerequisite for correct visualization and accurate calculation of some structural properties. periodic boundary imaging A trajectory processing method that separates the single combined trajectory written by a locally enhanced sampling run into 1 trajectory per copy of the duplicated region. Locally enhanced sampling simulates several copies of one part of a molecule at once, and their coordinates are interleaved in the output. Splitting them is required before each copy can be analysed or the copies averaged. (UNVERIFIED) LES trajectory splitting analysis A modelling method that changes the atomic coordinates or the composition of a molecular structure, for example by adding or deleting atoms, rotating or translating molecules, or building new polymer chains. These operations are the basic tools for building, editing and preparing molecular models for simulation. (UNVERIFIED) structure modification modeling An ion randomization method is a procedure for placing ions in a simulation box by randomly swapping their positions with existing solvent molecules. This is often done after an initial, simpler placement to ensure the ions are well-distributed throughout the solvent. This method helps to create a more disordered and realistic initial distribution of ions, avoiding potential artifacts from placing them in a regular grid. ion randomization analysis A system setup procedure that builds a larger system, called a supercell, by copying the original simulation unit cell along 1 or more directions. It is commonly used in materials science simulations to model bulk crystals or surfaces from a smaller repeating unit. It is the usual way to build a periodic system large enough to reduce artifacts caused by a small simulation box. (UNVERIFIED) cell replication analysis A modelling method that multiplies all the atomic coordinates of a system by a scaling factor, so that the whole system expands or contracts uniformly. It is used to adjust the density of a system, and some barostats control pressure by scaling the box and the coordinates within it in this way. (UNVERIFIED) coordinate scaling modeling A system setup procedure that gives atoms their initial speeds and directions at the start of a molecular dynamics simulation. The velocities are usually drawn at random from the Maxwell-Boltzmann distribution for the chosen starting temperature. This starts the thermal motion of the system before equilibration. (UNVERIFIED) velocity assignment analysis An analysis method that computes a value at every point of a 3-dimensional grid covering a region of a simulation. The values may be how often solvent is found at each point, or the electrostatic potential there. The result is a map that can be viewed as a surface or a cloud rather than read as a number. (UNVERIFIED) volumetric map generation analysis A procedure that removes a subset of atoms, typically solvent and ions, from a molecular topology file. It is most often used after a simulation to make smaller topology files for the receptor, the ligand and their complex, which MM/PBSA and MM/GBSA free energy calculations need. It is an essential step in preparing files for these end-point free energy methods. (UNVERIFIED) topology stripping analysis A force field parameterization method that modifies or adds to an existing set of force field parameters without changing the main parameter files. It is usually done by reading a small extra file of new or revised parameters, which override the default values loaded from the force field library. It is the usual way to add parameters for a non-standard ligand or to apply a specific correction. (UNVERIFIED) force field parameter update modeling An average structure calculation is a structural analysis method that computes the average coordinates of all atoms over a simulation trajectory. This provides a single, representative structure that reflects the central conformation of the system during the simulation. It is important to note that this average structure may not be a physically realistic conformation itself, especially if the molecule is highly flexible. average structure calculation analysis Coordinate centering is a structural analysis method, often performed as a pre-processing step, that translates the coordinates of a molecular system so that its center of mass is at the origin of the coordinate system. This is a standard procedure to ensure that the molecule remains in the center of the simulation box for easier analysis and visualization. coordinate centering analysis An analysis method that identifies, in each frame of a trajectory, the solvent molecules nearest to a chosen solute or region. It is used to reduce a trajectory to the solute plus a fixed number of nearest solvent molecules, and to record which solvent molecules were nearest in each frame, for example to estimate how long a water stays in contact. It answers which solvent molecules are nearest, rather than how much solvent there is. (UNVERIFIED) closest solvent analysis Trajectory frame filtering is a data processing step where a simulation trajectory is selectively reduced in size by removing certain frames based on a specific criterion. This could involve discarding an initial equilibration period or keeping only those frames where a molecule adopts a particular shape. This filtering is a common procedure used to focus an analysis on a specific time period or a subset of conformations that are most relevant to the scientific question. trajectory frame filtering A type of analysis method focused on examining the electronic properties of a molecule derived from its quantum mechanical description, such as the wavefunction or electron density. This includes analyzing the distribution of electrons (charge, spin density), the nature of molecular orbitals, chemical bonding characteristics, and electronic response properties like polarizability. It provides fundamental insights into chemical reactivity, spectroscopy, and intermolecular forces. (UNVERIFIED) electronic structure analysis An electronic structure analysis that assigns partial charges to the individual atoms of a molecule, either by dividing up its electron density or by fitting charges to its electrostatic potential. Many schemes exist, such as Mulliken, Löwdin, Hirshfeld, natural population analysis and electrostatic potential fitting, and they can give quite different values for the same molecule. The charges are used to parameterize classical force fields and to interpret polarity and reactivity. (UNVERIFIED) charge analysis An electronic structure analysis method that directly calculates molecular properties as expectation values using the computed quantum mechanical wavefunction or density matrix. This can include properties like dipole and multipole moments, electron density at specific points (e.g., bond critical points in QTAIM), spin densities, or electric field gradients, providing detailed information about the molecule's electronic state and interactions. It leverages the full information contained within the wavefunction. (UNVERIFIED) wavefunction property analysis A form of electronic structure analysis that investigates the molecular orbitals (MOs) of a system, which represent the energy levels and spatial distributions of electrons within the molecule based on quantum mechanics. Common analyses include visualizing frontier orbitals (HOMO, LUMO), examining orbital compositions (e.g., atomic orbital contributions), analyzing orbital energies for insights into ionization potential or electron affinity, and relating MOs to chemical bonding, reactivity, and electronic transitions. It helps rationalize chemical behavior based on electron distribution in orbitals. (UNVERIFIED) orbital analysis An electronic structure analysis technique that examines how the electron distribution of a molecule is distorted (polarized) in response to an external electric field or the field generated by nearby molecules. This involves calculating the induced dipole moment and the polarizability tensor, which quantify the molecule's susceptibility to charge redistribution. Understanding polarization is crucial for accurately describing intermolecular interactions, solvation effects, and responses to electromagnetic fields. (UNVERIFIED) polarization analysis A system setup procedure that puts the atoms of a system into the order that simulation and analysis programs expect, for example with the atoms of each molecule numbered consecutively, or matching the order in a topology file. Atoms in the wrong order are a common source of errors when preparing a system for simulation. Fixing the order ensures that coordinates and parameters are matched correctly for every atom. (UNVERIFIED) atom order fixing analysis OBSOLETE. Image bond fixing is a post-processing step applied to simulation trajectories to correct the visual representation of molecules in a periodic system. When a molecule crosses a periodic boundary, it can appear "broken" in the raw output, with parts of it on opposite sides of the simulation box. This procedure identifies such cases and translates the atomic coordinates to make the molecule appear whole, which is essential for correct visualization and many types of analysis. Retired as redundant with its former parent 'periodic boundary imaging' (MOLSIM:000659), of which it was the only child. Bond fixing is an implicit outcome of imaging, and that term's definition now states so. Reported and approved by the term verifier in GitHub issue #45. obsolete image bond fixing analysis true An atom selection method is the procedure or syntax used to specify a particular group of atoms within a molecular system for an analysis or modification. This is typically done using a text-based selection language that allows users to choose atoms based on properties like name, residue, or position. atom selection analysis A polarizable force field that represents charge as smeared-out Gaussian distributions rather than as points, with both permanent and induced multipole terms. Smearing the charge removes the artificial divergence that point charges give at very short range. It is markedly more accurate and more expensive than a point-charge treatment. (UNVERIFIED) GEM force field Gaussian electrostatic model force field Amber ff19SB is a modern and widely used set of rules, or a force field, for accurately simulating the structure and movement of proteins. The AMBER ff19SB protein force field is a recent refinement of the ff14SB lineage, incorporating updated quantum mechanical calculations for backbone dihedral angle parameters to improve the description of protein secondary structure balance. It is considered a state-of-the-art choice for all-atom protein simulations, offering high fidelity for studying protein folding, dynamics, and interactions. AMBER ff19SB A structural analysis method that measures the direction of vectors defined within a molecule, such as a bond vector or the normal to a plane, and how those directions change. It covers both the orientation of a single vector over time and the angle between 2 of them. It is how the orientation and rotational behaviour of specific groups is quantified. (UNVERIFIED) vector analysis An atom remapping method is a procedure used to systematically change the properties of atoms in a topology file, such as their atom types or charges. This is often required when converting a system from one force field to another or when applying specific corrections. For experts, this is an advanced manipulation used in force field development or for setting up non-standard simulations. (UNVERIFIED) atom remapping analysis A quantum chemistry method that solves the electronic Schrödinger equation using only fundamental physical constants and the identities of the atoms, with no parameters fitted to experiment. Hartree-Fock and the correlated methods built on it, such as Møller-Plesset perturbation theory and coupled cluster, belong here. The absence of fitted parameters is what lets accuracy be improved systematically rather than by re-tuning. (UNVERIFIED) ab-initio method A PDB ID is the unique four-character alphanumeric code (e.g., '1TIM', '6M0J') where the first character is always numeric. Every PDB id is assigned to a experimentally determined biomolecular structure deposited in the Protein Data Bank (PDB). This identifier serves to reference and retrieve PDB data. However these identifiers are now deprecated in favor of the Extended PDB ID. PDB ID A UniProt ID, more formally known as a UniProt Accession Number (e.g., 'P00761', 'Q8WZ42'), is the stable and unique identifier assigned to a protein sequence record in the UniProt Knowledgebase (UniProtKB). This ID provides access to comprehensive information about a protein, including its sequence, function, annotations, and cross-references to other databases. It serves as a primary key for protein sequence and functional data. UniProt ID Any origin from which data used in a biomolecular study is obtained, whether a database, a document, a laboratory record or a person. It covers where a structure, a parameter or a measurement came from. Recording it is what lets a claim in a dataset be traced back and checked. (UNVERIFIED) How this branch is built. The classes here are KINDS of data source, for example protein structural database, pathway database, or simulation data repository. Each real, named resource is then added as an individual of one of those classes, not as a class of its own. PDB, UniProt, ChEMBL, Reactome, MoDEL and AlphaFold DB are individuals in this way. This is on purpose: a named database is one single thing in the world, not a kind of thing, so it fits better as an individual. Please keep this pattern, and do not turn these individuals into classes as a tidy-up. Two things to know. First, individuals are not copied into the OBO-format file (see the release-format note in the ontology header). Second, most OBO ontologies use classes for everything, so a reviewer may ask why we did this. TODO:may be be replacable with IAO's information content entity (http://purl.obolibrary.org/obo/IAO_0000030 ) as an import. discussion needed: the definition there is however rather abstract. #ontologyengineer #import data source A data source that is an organized, usually electronic, collection of structured data that can be searched, retrieved and updated. In the biomolecular sciences such collections hold molecular structures, sequences, experimental results and simulation trajectories. They serve as shared resources for the research community. (UNVERIFIED) database source A type of database that archives original data directly submitted by researchers or generated through primary experimental efforts. These databases are foundational resources, providing the raw material (e.g., atomic coordinates, gene sequences) upon which further analysis and knowledge discovery are built. Examples include repositories for protein structures or DNA sequences. primary database A primary database specializing in the storage and dissemination of three-dimensional (3D) atomic coordinate data for biological macromolecules like proteins and nucleic acids. These databases are essential for understanding molecular function, designing drugs, and providing starting points for simulations. The Protein Data Bank (PDB) is a prime example. structural database A structural database specifically focused on archiving and distributing 3D atomic coordinates and associated experimental data for proteins and their complexes. These resources are critical for researchers studying protein architecture, function, and interactions. The PDB is the foremost example in this category. (UNVERIFIED) protein structural database A structural database dedicated to collecting and providing 3D atomic coordinates and related information for nucleic acids such as DNA and RNA, including their various forms and complexes. These databases support research into gene regulation, RNA function, and nucleic acid-drug interactions. The Nucleic Acid Database (NDB) is a key example. nucleic acid structural database A structural database that archives 3D maps and associated metadata derived from electron microscopy experiments, particularly cryo-electron microscopy (cryo-EM), which is used to determine the structure of large biomolecular assemblies. It complements X-ray crystallography and NMR databases by providing structural information for systems often intractable by other methods. The Electron Microscopy Data Bank (EMDB) is the main repository. (UNVERIFIED) electron microscopy database A primary database that stores and organizes linear sequences of amino acids (for proteins) or nucleotides (for DNA/RNA). These databases are fundamental for comparative genomics, evolutionary studies, identifying functional motifs, and linking sequence to structure and function. Examples include GenBank for nucleic acids and UniProtKB for protein sequences. sequence database A sequence database specifically focused on collecting, annotating, and distributing protein amino acid sequences and associated functional information. These resources are vital for understanding protein evolution, function, and for identifying proteins from experimental data. UniProtKB is a comprehensive example. protein sequence database A sequence database dedicated to archiving and providing access to DNA and RNA nucleotide sequences from various organisms, including genes, genomes, and transcripts. These databases are cornerstones of genomics and molecular biology research. GenBank, ENA, and DDBJ are major international examples. nucleic acid sequence database A primary database that curates information about small chemical compounds, including their structures, properties, and biological activities. These resources are crucial for drug discovery, cheminformatics, and understanding ligand-protein interactions. PubChem and ChEBI are prominent examples. (UNVERIFIED) small molecule database A small molecule database primarily focused on storing and providing access to 2D and 3D chemical structures, identifiers, and physicochemical properties of diverse chemical compounds. These are fundamental for chemical information retrieval and computational chemistry. PubChem Compound and ZINC are key examples. (UNVERIFIED) chemical structure database A small molecule database specifically curating information about drugs, including their chemical properties, pharmacological actions, targets, and therapeutic uses. These resources are vital for pharmacology, drug development, and clinical research. DrugBank is a well-known example. (UNVERIFIED) drug information database A primary database that archives experimental data obtained from various spectroscopic techniques, such as Nuclear Magnetic Resonance (NMR), mass spectrometry, or infrared spectroscopy, applied to biological molecules. These databases allow for the reanalysis of raw or processed spectral data and aid in structural and functional characterization. The Biological Magnetic Resonance Data Bank (BMRB) is an example for NMR data. spectroscopic database A spectroscopic database specifically designed to store, organize, and disseminate experimental data from Nuclear Magnetic Resonance (NMR) studies of biomolecules. This includes raw spectral data, assigned chemical shifts, coupling constants, and derived structural constraints, complementing 3D coordinate depositions. The BMRB is the principal repository. NMR data repository A type of database that derives its content from analyzing, curating, or integrating information from one or more primary databases, often providing value-added annotations or classifications. These databases help researchers interpret primary data by, for example, grouping proteins into families or classifying structures. Pfam (protein families) and SCOP (structural classification) are examples. secondary database A secondary database that classifies proteins into families and identifies conserved domains based on sequence similarity, structural features, or functional annotations. These resources help predict protein function and understand evolutionary relationships. Examples include Pfam, InterPro, and SMART. (UNVERIFIED) protein family and domain database A secondary database that organizes known protein structures into a hierarchical classification based on their evolutionary relationships and structural similarities, such as folds and superfamilies. These classifications provide insights into protein architecture and evolution. SCOP and CATH are prominent examples. (UNVERIFIED) structural classification database Databases that curate and represent information about biological pathways, such as metabolic, signaling, or regulatory networks, detailing the interactions between molecules. These resources help understand complex biological processes and the roles of individual components. KEGG and Reactome are widely used examples. pathway database A secondary database that collects and organizes experimental data on the biological activity of chemical compounds, particularly their interactions with protein targets, such as binding affinities and inhibition constants. These are crucial for drug discovery and chemical biology research. ChEMBL and BindingDB are key examples. (UNVERIFIED) bioactivity database A specialized type of database designed to store complex, structured, and unstructured information, often integrating data from multiple sources and using ontologies or controlled vocabularies to represent knowledge explicitly. Knowledge bases aim to provide a comprehensive understanding of specific biological entities or systems, facilitating inference and discovery. UniProtKB is a prime example in the protein domain. (UNVERIFIED) knowledge-based database A type of database primarily focused on the storage, preservation, and dissemination of datasets, often associated with specific types of studies, computational outputs, or large-scale projects. Repositories ensure data accessibility and reusability for the scientific community. Examples include those for simulation trajectories or predicted protein structures. repository database A repository specifically designed to store, share, and allow re-analysis of data generated from biomolecular simulations, such as molecular dynamics trajectories and analysis results. These resources promote reproducibility and enable new discoveries from existing simulation data. MDposit and MDrepo are examples. simulation data repository A repository that archives and provides access to computationally predicted three-dimensional structures of proteins, often generated by large-scale initiatives or community efforts. These resources offer structural hypotheses for proteins lacking experimental structures. The AlphaFold DB is a prominent example. predicted structure repository A repository dedicated to storing and disseminating datasets from metabolomics experiments, which involve the comprehensive study of small molecules (metabolites) within cells, tissues, or organisms. These resources support systems biology and biomarker discovery. MetaboLights is an example. (UNVERIFIED) metabolomics repository A database that indexes and provides access to bibliographic information from scientific journals, conference proceedings, and other scholarly publications. These are essential tools for researchers to find relevant scientific articles and stay updated on research findings. PubMed and Scopus are widely used examples. literature database A web server that runs a computation on data the user supplies, rather than returning records it already holds. The user uploads a structure, the server computes something from it, and the result comes back. That is what separates it from a database server, which retrieves rather than computes. (UNVERIFIED) tool web server A data format that holds the direct output of a prediction program, such as a predicted structure, a predicted binding site, or a predicted functional annotation. It covers the result as the program wrote it, before any curation or deposition into a public repository. A coordinate file from a local homology modelling run is an example. (UNVERIFIED) prediction data format A visualization and analysis program for molecular structures, built to handle very large data such as cryogenic electron microscopy maps and whole virus capsids. It is the successor to Chimera, rewritten for modern graphics hardware. Its speed on large structures is what distinguishes it. (UNVERIFIED) ChimeraX UCSF ChimeraX OBSOLETE. Refers to data obtained directly from a specific laboratory experiment. This can include data reported in a scientific publication for which no separate database entry exists, or unpublished results from ongoing research. Retired 2026-09-18. MOLSIM:000688 data source sorted its children by 3 different questions at once: what kind of resource it is, how the data was produced, and who may see it. Only the first is a kind of source, so a private unpublished PDB entry belonged under all 3. The 2 facets are now recorded by the MOLSIM:002274 data source descriptor branch instead. Raised in review by orcid:0000-0002-2294-5393; neither this term nor its former children had a single instance. obsolete experimental data source true OBSOLETE. A published research article, review, book chapter, or conference proceeding that serves as the source of specific data, findings, or methodologies. When data is cited directly from a publication without an associated database entry, the publication itself is considered the data source. This is common for specific measurements or observations not typically archived elsewhere. Retired 2026-09-18 in favour of IAO:0000310 document, which is now imported and attached under MOLSIM:000688 data source. The term duplicated a concept IAO already models, and its own definition enumerated exactly IAO's document types. Part of the data source refactor asked for in review by orcid:0000-0002-2294-5393, which removed the facets experimental and private from the hierarchy and left it sorting sources by what kind of resource they are. obsolete scientific publication source true OBSOLETE. Data or findings originating from a laboratory experiment that have not yet been published or formally disseminated to the public or scientific community. This category covers preliminary results, ongoing research data, or information shared through informal channels before formal publication or deposition. Retired 2026-09-18. MOLSIM:000688 data source sorted its children by 3 different questions at once: what kind of resource it is, how the data was produced, and who may see it. Only the first is a kind of source, so a private unpublished PDB entry belonged under all 3. The 2 facets are now recorded by the MOLSIM:002274 data source descriptor branch instead. Raised in review by orcid:0000-0002-2294-5393; neither this term nor its former children had a single instance. obsolete unpublished experimental result source true OBSOLETE. A data source that is not publicly accessible, such as an internal database maintained by a private company or research institution, or data held privately by an individual researcher. This category is important for acknowledging data origins that cannot be independently verified or accessed by the wider community. Retired 2026-09-18. MOLSIM:000688 data source sorted its children by 3 different questions at once: what kind of resource it is, how the data was produced, and who may see it. Only the first is a kind of source, so a private unpublished PDB entry belonged under all 3. The 2 facets are now recorded by the MOLSIM:002274 data source descriptor branch instead. Raised in review by orcid:0000-0002-2294-5393; neither this term nor its former children had a single instance. obsolete private source true OBSOLETE. A database developed and maintained within a specific organization (e.g., a pharmaceutical company, a research institute) for its internal use, not generally accessible to the public. These often contain proprietary data, specialized collections, or information tailored to the organization's specific research needs. Retired 2026-09-18. MOLSIM:000688 data source sorted its children by 3 different questions at once: what kind of resource it is, how the data was produced, and who may see it. Only the first is a kind of source, so a private unpublished PDB entry belonged under all 3. The 2 facets are now recorded by the MOLSIM:002274 data source descriptor branch instead. Raised in review by orcid:0000-0002-2294-5393; neither this term nor its former children had a single instance. This one has a successor: an in house database is a database, so MOLSIM:000689 database source replaces it. obsolete in house database true Information or data obtained directly from an individual researcher or expert through informal means such as email, conversation, or unpublished notes, rather than through a formal publication or public database. While sometimes necessary, such sources are typically less verifiable and are often used for supplementary or preliminary information. private communication source personal communication source A type of force field classified by the level of structural detail, or resolution, used to represent the system's components. This classification distinguishes between models that represent every atom explicitly versus those that group atoms into larger pseudo-atoms or beads. The resolution impacts the balance between computational efficiency and physical accuracy. (UNVERIFIED) resolution-based force field A force field in which groups of atoms are represented by single interaction sites rather than individually. Cutting the number of particles cuts the cost of every step and allows a longer timestep, so much larger systems and much longer times become reachable. What is lost is any detail below the size of a site. (UNVERIFIED) coarse-grained force field A force field that explicitly includes parameters for every atom in the simulated system, including hydrogen atoms. This approach provides a high level of detail and aims for greater physical realism compared to coarse-grained models. All-atom simulations are generally more computationally intensive due to the larger number of particles. all-atom force field A force field for carbohydrates (glycans) in the AMBER family, developed by the GLYCAM group, that provides parameters for common monosaccharides and the glycosidic linkages between them. It is used to study oligosaccharides and the carbohydrate portions of glycoproteins and glycolipids. It is designed to be compatible with AMBER protein and lipid force fields. (UNVERIFIED) AMBER GLYCAM 06 A force field for carbohydrates in the AMBER family, a variant of GLYCAM06, in which extra points representing lone pairs are placed on the oxygen atoms so that it can be used with the 5-site TIP5P water model. The distance of each extra point from its oxygen was chosen to best reproduce the quantum electrostatic potential, and the fit leaves almost no charge on the oxygen nucleus itself. The extra points make hydrogen bonding more directional, which matters for sugars, whose behaviour depends strongly on hydrogen bonds. (UNVERIFIED) AMBER GLYCAM 06 EP A set of ion parameters for the AMOEBA polarizable force field as distributed with the AMBER suite. Because AMOEBA represents charge with atomic multipoles and explicit induced dipoles rather than fixed point charges, ions need parameters fitted for that form. Parameters from a fixed-charge force field cannot be used with it. (UNVERIFIED) AMBER AMOEBA A specific protein force field parameter set from the AMBER family, representing one of the earlier polarizable force fields. Unlike fixed-charge force fields (like ff99SB), ff02 includes electronic polarizability explicitly, aiming for a more accurate description of electrostatics but at a higher computational cost. Its adoption was less widespread than fixed-charge counterparts. AMBER ff02 A force field from the AMBER family, introduced in 2010, that combines the ff99SB protein parameters with the bsc0 correction for DNA and the chi OL3 correction for RNA. It brought together the recommended protein and nucleic acid parameters of that time in 1 set. It was later superseded by ff12SB and ff14SB for proteins and by newer nucleic acid force fields. (UNVERIFIED) AMBER ff10 A specific protein force field parameter set from the AMBER family, representing an update released around 2012. Compared to ff99SB, ff12SB refined quantum-chemical (QM) based side-chain dihedral parameters and updated the backbone φ rotational profile to compensate for biases seen in earlier version. It served as an intermediate improvement before the release of ff14SB. AMBER ff12SB An all-atom force field for proteins and nucleic acids from the AMBER family, described by Cornell and co-workers in 1995, whose partial charges were fitted to quantum electrostatic potentials with the RESP method. Later AMBER sets such as ff96, ff99 and ff99SB revised its parameters, mainly the backbone torsion terms. It gives a poor balance between helical and extended protein structures and is now obsolete. (UNVERIFIED) AMBER ff94 A protein force field in the AMBER family that revises the ff94 parameters, mainly the backbone torsion terms, to reduce the preference of ff94 for helical structure. It has been replaced by later AMBER protein force fields. (UNVERIFIED) AMBER ff96 A force field for proteins and nucleic acids in the AMBER family, described by Wang, Cieplak and Kollman in 2000, that revises many torsion and other bonded parameters of ff94 against quantum calculations on small model molecules. It is the base of most later AMBER force fields, such as ff99SB for proteins and the bsc0 and OL corrections for nucleic acids. Its backbone torsions still gave a poor balance between helical and extended protein structure, which ff99SB corrected. (UNVERIFIED) AMBER ff99 A protein force field, introduced by Sorin and Pande in 2005, that changes the backbone phi torsion terms of AMBER ff99 to reduce its overly strong preference for helical structure. It was developed to study how helical peptides fold, and it agreed better than ff99 with measured folding rates and stabilities. (UNVERIFIED) AMBER ff99 phi A force field for nucleic acids, built on the AMBER ff99 parameter set with the alpha and gamma backbone dihedral terms refit by the Barcelona group. The refit removed the build-up of non-native gamma trans backbone geometries that made DNA simulations degrade on the 50 to 100 nanosecond timescale. It became the base of later AMBER DNA and RNA force fields, such as OL15 and OL3. (UNVERIFIED) AMBER ff99bsc0 A protein force field in the AMBER family that refits the backbone torsion terms of ff99 against high-level quantum calculations on glycine and alanine tetrapeptides. It gives a much better balance between helical and extended structures and corrects the sampling of glycine. The SB in its name refers to Stony Brook, where it was developed, and it is the base of later variants such as ff99SB-ILDN and ff14SB. (UNVERIFIED) AMBER ff99SB A force field for proteins, built on the AMBER ff99SB parameter set with the side-chain torsion terms refit to high-level quantum calculations for isoleucine, leucine, aspartate and asparagine, whose 1-letter codes give the name. These 4 residues were chosen because their simulated rotamer populations differed most from Protein Data Bank statistics. The new terms agree considerably better with NMR measurements of side-chain conformations, and the rest of the parameter set is unchanged. (UNVERIFIED) AMBER ff99SB ILDN An all-atom force field parameter set from the CHARMM family, widely used for protein simulations, often paired with the CMAP (Cross-term energy Correction Map) correction for backbone torsions. Released before CHARMM27 and CHARMM36, it represented a standard for many years. Parameterization relied heavily on fitting to experimental data available at the time. (UNVERIFIED) CHARMM 22 An earlier force field parameter set from the CHARMM family, notable for being a 'united-atom' force field for proteins. In this representation, nonpolar hydrogen atoms (e.g., on CH, CH2, CH3 groups) are implicitly included in the heavy atoms they are attached to, reducing the number of particles and computational cost compared to all-atom models. CHARMM 19 An all-atom force field of the CHARMM family that combines the CHARMM22 protein parameters, usually with the CMAP backbone correction, with nucleic acid parameters published in 2000. In GROMACS the name refers to CHARMM22 plus CMAP for proteins. It was later replaced by CHARMM36. (UNVERIFIED) CHARMM 27 A protein force field that is version 1 of OPEP, a coarse-grained model representing each amino acid by its backbone atoms N, H, Cα, C and O and a single bead for the side chain. Keeping the backbone atoms retains much of the detail of backbone hydrogen bonding while reducing computational cost. (UNVERIFIED) OPEP 1 A protein force field that is version 2 of OPEP, a coarse-grained model representing each amino acid by its backbone atoms N, H, Cα, C and O and a single bead for the side chain. Keeping the backbone atoms retains much of the detail of backbone hydrogen bonding while reducing computational cost. (UNVERIFIED) OPEP 2 A protein force field that is version 3 of OPEP, a coarse-grained model representing each amino acid by its backbone atoms N, H, Cα, C and O and a single bead for the side chain. Keeping the backbone atoms retains much of the detail of backbone hydrogen bonding while reducing computational cost. (UNVERIFIED) OPEP 3 A protein force field that is version 4 of OPEP, a coarse-grained model representing each amino acid by its backbone atoms N, H, Cα, C and O and a single bead for the side chain. Keeping the backbone atoms retains much of the detail of backbone hydrogen bonding while reducing computational cost. (UNVERIFIED) OPEP 4 A family of force fields developed by Jorgensen and co-workers for condensed-phase simulations, whose non-bonded parameters were fitted mainly to reproduce measured properties of organic liquids, such as density and heat of vaporisation. It includes united-atom and all-atom versions and covers organic molecules and proteins, with OPLS-AA/M the latest version for proteins. (UNVERIFIED) OPLS A force field of the OPLS family in which every atom, including every hydrogen, is represented as a separate interaction site. Representing hydrogens explicitly describes groups such as aromatic rings and hydrogen-bonding sites in more detail than the united-atom version, at a higher cost. Its parameters exist for proteins and many organic molecules. (UNVERIFIED) OPLS-AA A force field of the OPLS family in which non-polar hydrogens are merged into the heavy atom they are bonded to, so that a methyl group is 1 interaction site rather than 4. Merging them cuts the particle count substantially at the cost of any detail those hydrogens carried. It was widely used for organic liquids before all-atom parameters became affordable. (UNVERIFIED) OPLS-UA A specific parameter set version within the OPLS-AA force field family, released around 2001. It contained parameters primarily for proteins and organic liquids available at that time. Subsequent versions like OPLS-2005 updated and expanded upon this set. (UNVERIFIED) OPLS-2001 A significant update and expansion of the OPLS-AA force field parameter set, released around 2005. It included refined parameters for proteins, nucleic acids, and a wider range of organic functional groups compared to OPLS-2001. It remains a widely used parameter set within the OPLS family. (UNVERIFIED) OPLS-2005 A rigid water model with 4 sites, the fourth being a massless point near the oxygen, on the bisector of the H-O-H angle, carrying that atom's negative charge. Moving the negative charge off the oxygen towards the hydrogens gives a better electrostatic field around the molecule. It costs more than a 3-site model and reproduces properties such as the dielectric constant and the temperature of maximum density better than TIP3P. (UNVERIFIED) TIP4P A specific variant of the TIP4P water model, optimized for use with Ewald summation methods for calculating long-range electrostatic interactions, which are standard in modern simulations with periodic boundary conditions. The parameters are adjusted to provide good agreement with experimental properties of liquid water (like density, enthalpy of vaporization) under typical simulation conditions (room temperature and pressure). It is a popular choice for biomolecular simulations. (UNVERIFIED) TIP4P-Ew A variant of the TIP4P water model, parameterized to reproduce the melting temperature of ordinary ice and the properties of the different ice phases. It gives a better description of water's phase diagram, particularly the solid phases, than models optimized for the liquid state. It is useful for studying freezing, melting, or ice interfaces. (UNVERIFIED) TIP4P/Ice A variant of the TIP4P water model, published by Abascal and Vega in 2005, re-parameterized to provide improved agreement with a wider range of experimental properties of liquid water over various temperatures and pressures compared to earlier TIP4P versions. It is considered one of the most accurate rigid, non-polarizable 4-point models for general liquid water simulations. It accurately predicts properties like the temperature of maximum density. (UNVERIFIED) TIP4P/2005 A rigid water model with 5 sites, adding 2 massless points in the positions the oxygen's lone pairs would occupy. Placing negative charge there makes hydrogen bonding directional in the way real water's is. It was fitted to reproduce the temperature of maximum density of water near 4 °C, at a higher cost than 3-site and 4-site models. (UNVERIFIED) TIP5P A rigid water model with 6 sites, carrying charges on all 3 atoms, on 2 lone-pair points and on a further point near the oxygen. The extra sites refine the electrostatic field further still. Its cost restricts it to studies of water itself rather than of biomolecules in water. (UNVERIFIED) TIP6P A rigid water model with 3 interaction sites, 1 on each atom, an O-H bond length of 1 Å and an H-O-H angle of 109.47 degrees, the ideal tetrahedral value, rather than the angle of an isolated molecule. Like other fixed-charge models, its charges give a dipole moment larger than that of an isolated water molecule, to represent on average the polarization of water in the liquid. It is among the oldest water models still in use, and SPC/E was derived from it. (UNVERIFIED) SPC A variation of the SPC water model where the normally rigid constraints on bond lengths (O-H) and/or angle (H-O-H) are replaced with flexible harmonic potentials. This allows the water molecule geometry to fluctuate during the simulation, which can affect dynamics and spectroscopic properties. It adds computational overhead compared to rigid SPC. (UNVERIFIED) SPC flexible A rigid 3-site water model that keeps the geometry of its predecessor and adds a correction for the energy cost of polarizing a molecule from its gas-phase to its liquid charge distribution. That correction improves the density and the diffusion coefficient markedly. It is a common choice where bulk water properties matter more than compatibility with a particular force field. (UNVERIFIED) SPC/E A force field specifically developed and parameterized to accurately model a particular class of chemical molecules, such as proteins, lipids, or nucleic acids. Its parameters are tuned based on experimental data and/or quantum mechanical calculations relevant to that molecular type. This specialization aims to improve the accuracy of simulations for specific biomolecular systems. (UNVERIFIED) chemical entity-based force field A quantum chemistry engine built around density functional theory using Slater-type orbitals rather than the Gaussian functions most programs use. Slater functions have the correct shape at the nucleus, which suits properties that depend on the region near heavy atoms. It is often chosen for transition metal and relativistic calculations. (UNVERIFIED) ADF Abalone (Abalone II) is a general purpose molecular dynamics and molecular graphics program for simulations of bio-molecules in a periodic boundary conditions in explicit or implicit water models. Mainly designed to simulate the protein folding and DNA-ligand complexes in AMBER force field. Abalone An open-source quantum chemistry engine for atomistic simulation of solids, liquids, molecules and biological systems, with a focus on density functional theory. Its Quickstep module uses a mixed basis of Gaussian functions and plane waves, which makes density functional calculations on large systems efficient, including ab initio molecular dynamics and QM/MM simulations that treat a small region with quantum mechanics and the rest classically. It is written to run in parallel on large computers. (UNVERIFIED) CP2K A quantum chemistry engine for ab initio molecular dynamics, in which the forces on the atoms come from density functional theory, usually with a plane wave basis. Its main method, Car-Parrinello molecular dynamics, which gives the program its name, lets the electronic orbitals move with their own fictitious dynamics so that the electronic structure need not be fully solved again at every step. It also performs Born-Oppenheimer molecular dynamics. (UNVERIFIED) CPMD A high-performance software package specifically designed for classical molecular dynamics simulations of biological systems, developed by D. E. Shaw Research (DESRES). Desmond is known for its exceptional speed, particularly on GPUs, and its integration with the Schrödinger Materials Science Suite. It supports standard force fields (like CHARMM, AMBER, OPLS) and offers features for system building, simulation execution (including free energy calculations), and analysis, primarily targeting drug discovery and biomolecular research. Desmond A quantum chemistry engine offering a broad range of wavefunction and density functional methods, heavily optimised for Intel processors. It shares its origins and much of its input syntax with GAMESS (US). It was formerly distributed under the name PC GAMESS. (UNVERIFIED) Firefly A quantum chemistry engine of the GAMESS family, developed in the United Kingdom separately from GAMESS-US although the 2 share common origins. It offers a broad range of ab initio and density functional methods for the electronic structure of molecules, and can be combined with molecular mechanics in QM/MM calculations, where a small region is treated with quantum mechanics and the rest classically. Like other quantum chemistry programs, it is used mainly for electronic structure calculations and parameterization rather than for large-scale classical molecular dynamics. (UNVERIFIED) GAMESS-UK A quantum chemistry engine of the GAMESS family, developed in the United States and freely available, that provides methods from semi-empirical and Hartree-Fock theory through density functional theory to perturbation, coupled cluster and multi-configuration methods. It also performs geometry optimization, vibrational analysis and QM/MM calculations, which treat a small region with quantum mechanics and the rest classically. It is widely used in computational chemistry research and teaching. (UNVERIFIED) GAMESS-US A software package for classical molecular dynamics simulation, developed principally at ETH Zürich, covering system setup, simulation and analysis. It is distributed together with the force field family of the same name, which is a separate thing from the program. (UNVERIFIED) GROMOS A commercial molecular modeling suite for desktop computers with a graphical interface for building, viewing and simulating molecules. It combines molecular mechanics with several force fields, semi-empirical and ab initio quantum chemistry, and molecular dynamics in 1 program. (UNVERIFIED) HyperChem A commercial quantum chemistry engine from Schrödinger, used mainly for density functional and Hartree-Fock calculations, that evaluates the costly electron repulsion terms quickly with a pseudospectral method. It also offers local second-order Møller-Plesset perturbation theory (LMP2) and QM/MM calculations, which treat a small region with quantum mechanics and the rest classically. It is often used in drug discovery for reaction modeling, property prediction and force field parameterization. (UNVERIFIED) Jaguar A classical molecular dynamics engine, open source and developed at Sandia National Laboratories, that is designed to run in parallel on many processors and supports a very wide range of interaction potentials. Its applications reach from biomolecules, using force fields such as CHARMM, AMBER and OPLS, to materials science, solid-state physics and coarse-grained models. That breadth makes it suitable for simulation setups that other engines do not support. (UNVERIFIED) LAMMPS A semi-empirical quantum chemistry engine that implements methods such as AM1, PM3, PM6 and PM7, which replace most of the costly integrals of ab initio theory with parameters fitted to reference data. These methods are much faster than ab initio or density functional calculations but usually less accurate, so they allow larger systems or more calculations. It is often used for fast geometry optimization, property estimates and as the quantum part of QM/MM calculations where speed matters. (UNVERIFIED) MOPAC A highly scalable, parallel molecular dynamics simulation software package developed at the University of Illinois Urbana-Champaign, designed specifically for simulating large biomolecular systems on high-performance computers (NAnoscale Molecular Dynamics). NAMD works particularly well with the CHARMM force field (and others like AMBER) and is often used in conjunction with the VMD visualization program for setup and analysis. It is known for its excellent performance on parallel architectures, including GPUs. NAMD An open-source quantum chemistry engine for density functional theory calculations and ab initio molecular dynamics that uses numerical atomic orbitals, confined to a region around each atom, as its basis, together with pseudopotentials. That basis keeps calculations on large systems affordable and suits parallel computers. It is used mostly in materials science and condensed matter physics, although it can also be applied to biomolecules. (UNVERIFIED) SIESTA A commercial molecular modeling and computational chemistry suite with an integrated graphical interface for building, viewing and calculating molecules. It implements molecular mechanics, semi-empirical methods such as AM1 and PM3, Hartree-Fock, density functional theory and Møller-Plesset perturbation theory. It is widely used in teaching and in research on how molecular structure relates to properties. (UNVERIFIED) Spartan A freely available software package for molecular mechanics and molecular dynamics, best known for its implementation of polarizable force fields such as AMOEBA. It supports several standard force fields, including AMBER, CHARMM, OPLS and MMFF, and includes programs for energy minimization, molecular dynamics, free energy calculations and structural analysis. Its design is modular. (UNVERIFIED) TINKER A molecular dynamics engine for high-performance parallel computers, developed at Fraunhofer SCAI. It supports many interaction potentials and is used mainly in nanotechnology and materials science, and also in biochemistry and biophysics. (UNVERIFIED) Tremolo-X A quantum chemistry program package for molecules and clusters, built for speed and parallel performance, that offers density functional theory and correlated wavefunction methods such as second-order Møller-Plesset perturbation theory and coupled cluster theory. It is widely used where high accuracy or large basis sets are needed, for example in reaction mechanism studies, spectroscopy and force field parameterization. (UNVERIFIED) TURBOMOLE A commercially licensed quantum chemistry engine, developed at the University of Vienna, for density functional calculations on periodic systems using plane waves and the projector augmented wave method. Plane waves naturally describe an infinite repeating solid. It is one of the most widely used programs for solids, surfaces and interfaces, including ab initio molecular dynamics. (UNVERIFIED) VASP A software package primarily used for macromolecular structure determination based on experimental data, particularly X-ray crystallography and NMR spectroscopy data. It incorporates molecular mechanics (energy terms from force fields), molecular dynamics (often used for refinement or simulated annealing protocols), and terms representing the experimental restraints to refine structures consistent with both physics and experimental observations. While it uses MD concepts, its primary goal is structure determination rather than simulating equilibrium dynamics. X-PLOR A barostat algorithm that gives the box volume its own equation of motion, as in the Andersen barostat, and combines it with a Nosé-Hoover thermostat, following the constant-pressure equations of motion proposed by Hoover. The equations are deterministic, so no random numbers are involved. They were later shown not to sample the isothermal-isobaric ensemble exactly, a problem corrected in the later equations of Martyna, Tobias and Klein. (UNVERIFIED) Hoover barostat A constraint algorithm that solves all the distance constraint equations of a molecule together with Newton's method, instead of correcting one constraint at a time as SHAKE does. Each iteration costs more, but the solution converges much faster, so far fewer iterations are needed. It was designed for small molecules, where solving the equations together is cheap. (UNVERIFIED) M-SHAKE algorithm A constraint algorithm that, before the SHAKE iteration, transforms the constraint equations with a precomputed matrix, called a preconditioner, so that they become nearly independent of one another. The decoupled constraints then converge much faster, quadratically rather than linearly, which means the error shrinks far more quickly from one iteration to the next. It suits rigid or semi-rigid molecules with many coupled constraints. (UNVERIFIED) P-SHAKE algorithm An analytical (non-iterative) constraint algorithm specifically designed for rigid water models like TIP3P or SPC/E, where both bond lengths and the bond angle are fixed. Because the geometry is simple and fixed, the constraint equations for a water molecule can be solved analytically, making SETTLE significantly faster than iterative methods like SHAKE for constraining water. It is highly efficient for simulations with explicit water. (UNVERIFIED) SETTLE algorithm A constraint algorithm that extends the SHAKE approach to rigid bodies of 3 or more atoms, correcting their positions after an unconstrained step by a rotation of each rigid body as a whole. It gives the same trajectory as fully converged SHAKE, but is more efficient and more accurate for groups of 3 or more atoms. It is used to keep rigid molecules or rigid fragments at their fixed geometry during a simulation. (UNVERIFIED) SHAPE algorithm A constraint algorithm that holds bond lengths fixed while using information about the atoms' velocities as well as their positions to converge faster. Holding bonds fixed removes the fastest vibrations from a system, which is what allows a longer timestep. It is an extension of the SHAKE approach rather than a separate idea. (UNVERIFIED) WIGGLE algorithm A thermostat algorithm that adds an extra degree of freedom to the system, representing the heat bath, which exchanges energy with the atoms through its own equation of motion. The equations are deterministic and generate the canonical distribution provided the dynamics are ergodic, that is explore all accessible states, which is not always the case for small or stiff systems. In Nosé's original form, time is rescaled as well, which Hoover's later reformulation removes. (UNVERIFIED) Nose thermostat A thermostat algorithm that applies a Poincaré time transformation to the Nosé Hamiltonian, so that the extended system keeps a Hamiltonian form, as in the original Nosé thermostat, but evolves in real time, as in the Nosé-Hoover thermostat. Keeping the Hamiltonian form allows it to be integrated with symplectic, time-reversible methods that give good long-term stability. (UNVERIFIED) Nose-Poincare thermostat A temperature control algorithm that modifies simple weak coupling by adding a random term chosen so the resulting distribution of kinetic energies is the correct one. Weak coupling alone drives every fluctuation towards the mean, which suppresses the spread a real system has. The random term restores it at no extra cost. (UNVERIFIED) tcoupl=v-rescale V-rescale thermostat tcoupl=v-rescale A quantum chemistry method that simplifies the equations of an ab initio or density functional method by neglecting some terms and replacing others with parameters fitted to experimental or reference data. It is much faster than ab initio methods while still describing electrons explicitly, which molecular mechanics does not. It gives a balance between cost and accuracy for properties such as molecular geometry and energy. (UNVERIFIED) semi-empirical method A semi-empirical method that treats every valence electron explicitly while representing the inner electrons by a fixed core. Including all valence electrons is what lets it describe molecules of any element combination rather than only conjugated ones. It is the level nearly all modern semi-empirical methods work at. (UNVERIFIED) all-valence-electron restricted A semi-empirical quantum method that treats only the pi electrons of a conjugated system, holding the rest of the molecule fixed as a framework. Restricting the calculation to pi electrons made conjugated molecules computable long before general methods existed. It predicted the absorption spectra of dyes well and cannot describe anything outside the pi system. (UNVERIFIED) PPP A semi-empirical method that treats only the pi electrons explicitly, holding the sigma electrons and nuclei fixed as a core framework. It applies to planar conjugated molecules, where the pi electrons control properties such as ultraviolet and visible absorption. It cannot describe anything outside the pi system, so it is used mainly for qualitative studies of conjugated molecules and for teaching. (UNVERIFIED) pi-electron restricted A semi-empirical quantum method built on the NDDO approximation, adding extra Gaussian terms to the core repulsion so that hydrogen bonding is described better than its predecessor manages. Its parameters were adjusted by hand against measured heats of formation and geometries. It remains in wide use, particularly as the basis of fast charge assignment schemes. (UNVERIFIED) AM1 A semi-empirical all-valence-electron method, in its second parameterisation, that neglects the differential overlap, the product of 2 different atomic orbitals, in the electron repulsion integrals even when both orbitals are on the same atom. The severity of that approximation makes it very fast and correspondingly crude. It is of historical interest rather than in current use. (UNVERIFIED) CNDO/2 A semi-empirical all-valence-electron method that keeps the 1-centre exchange integrals between different orbitals on the same atom, which the earlier complete neglect (CNDO) scheme discards. Keeping them is what lets it describe electron spin distributions and the properties that depend on them. It is the level the MINDO and SINDO families refine. (UNVERIFIED) INDO A fundamental ab initio quantum chemistry method that approximates the solution to the electronic Schrödinger equation for a multi-electron system. It treats each electron as moving in an average field created by all other electrons, neglecting instantaneous electron correlation but including exact exchange interaction. It serves as a starting point for more sophisticated ab initio methods and is the foundation simplified by semi-empirical approaches. (UNVERIFIED) HF Hartree-Fock A Hartree-Fock method for systems with unpaired electrons that keeps electrons of opposite spin in the same spatial orbitals wherever they are paired. The constraint means the wavefunction remains an eigenfunction of total spin, with no spin contamination. It is less flexible than the unrestricted method and usually gives a slightly higher energy. (UNVERIFIED) ROHF restricted open-shell Hartree-Fock An iterative computational procedure that solves for a set of molecular orbitals by repeatedly recomputing the average field each electron feels from all the others until the orbitals stop changing. It is how the Hartree-Fock equations, and those of density functional theory, are solved in practice. A calculation that fails to reach that fixed point is said not to have converged. (UNVERIFIED) SCF self-consistent field A Hartree-Fock method that lets electrons of opposite spin occupy different spatial orbitals. The extra freedom describes unpaired electrons better than forcing paired orbitals does. The price is that the resulting wavefunction is not an eigenfunction of total spin, so it can be contaminated by states of higher multiplicity. (UNVERIFIED) UHF Unrestricted Hartree-Fock A family of semi-empirical all-valence-electron methods based on the INDO (intermediate neglect of differential overlap) approximation, with parameters fitted to reproduce the heats of formation of organic molecules. Successive versions are numbered, the best known being MINDO/3. They were later replaced by methods based on the NDDO approximation, such as MNDO and AM1. (UNVERIFIED) MINDO A semi-empirical quantum method built on the NDDO approximation, with parameters fitted to measured heats of formation, geometries and dipole moments. It described geometries and heats of formation far better than the earlier INDO-based methods. Its known failure is that it underestimates hydrogen bonding severely, which is what later methods such as AM1 set out to fix. (UNVERIFIED) MNDO A semi-empirical approximation scheme that sets to zero every electron repulsion integral containing the product of 2 atomic orbitals centred on different atoms, while keeping those in which each product comes from orbitals on the same atom. Dropping those integrals is what makes NDDO-based calculations fast, and the approximation is less severe than in the CNDO and INDO schemes. It underlies the MNDO, AM1 and PM families of methods. (UNVERIFIED) NDDO A modification of the PM3 and MNDO semi-empirical methods that adds a pairwise distance directed Gaussian function to the core repulsion between each pair of atoms, together with a new set of parameters. The corrections improve heats of formation markedly, particularly for organic molecules. It changes the parameters and the core repulsion rather than the underlying approximation. (UNVERIFIED) PDDG A semi-empirical quantum method built on the NDDO approximation, with all of its parameters fitted automatically to a large set of measured molecular properties. Treating every parameter as adjustable, including some that AM1 took from spectroscopic measurements, is what distinguishes it from AM1. (UNVERIFIED) PM3 A variant of the PM3 semi-empirical method that adds a molecular mechanics correction for the HCON (peptide) linkage, a torsion term that raises the too-low rotation barrier PM3 gives for amide bonds. Without it, semi-empirical methods tend to make the amide nitrogen pyramidal and the amide bond too easy to twist. (UNVERIFIED) PM3MM A semi-empirical quantum method built on the NDDO approximation, parameterised for most of the periodic table rather than for organic elements alone. That coverage is what distinguishes it from its predecessors, which handle few metals. It also revised the core repulsion terms, improving hydrogen bonding and the description of metal complexes. (UNVERIFIED) PM6 A semi-empirical quantum method built on the NDDO approximation, refitting the parameters of AM1 against a larger and more modern set of measurements. It uses the same functional form as its predecessor and differs only in the parameter values. It covers the elements most common in organic and biological molecules. (UNVERIFIED) RM1 A semi-empirical quantum method built on the NDDO approximation that computes the 2-electron repulsion integrals it keeps from a minimal basis of Gaussian functions, instead of approximating them by a multipole expansion as earlier NDDO methods do. The computed integrals are then scaled by empirical functions to reproduce experimental results. It is much less widely used than its contemporaries. (UNVERIFIED) SAM1 A semi-empirical all-valence-electron method based on the INDO (intermediate neglect of differential overlap) approximation, in which the atomic orbitals are first orthogonalised symmetrically. The orthogonalisation makes the approximations more internally consistent than in the unmodified INDO scheme. It was developed to improve on earlier INDO-based methods. (UNVERIFIED) SINDO A modification of the AM1 semi-empirical method specifically parameterized to handle lanthanide elements (rare earth metals). Standard AM1 parameters are not available or reliable for these heavy elements, so Sparkle/AM1 replaces the core description of the lanthanide ion with a point charge dressed by Gaussian functions, adjusted to reproduce structural data. It enables AM1-level calculations on complexes containing lanthanides. (UNVERIFIED) Sparkle/AM1 An all-valence-electron semi-empirical method likely based on the INDO or NDDO approximation schemes, developed by Zhidomirov and coworkers. It represents one of the many variations within the semi-empirical framework, likely with specific parameterizations or modifications to the integral approximations aimed at particular applications or accuracy improvements. Note: This method is less common than MNDO/AM1/PMx variants, and specific details might require consulting specialized literature. (UNVERIFIED) ZANDO A post-Hartree-Fock method that recovers electron correlation by treating the difference between the true electronic Hamiltonian and the Hartree-Fock one as a perturbation, and computing energy corrections order by order. Each successive order adds more correlation and costs more. The order used is written as a number after the name. (UNVERIFIED) MP Moeller-Plesset A Møller-Plesset method truncated at second order, which adds the correlation energy from double excitations out of the Hartree-Fock reference. It is the cheapest wavefunction method that includes electron correlation and is widely used for that reason. It describes dispersion reasonably and overestimates it for large systems. (UNVERIFIED) MP2 Moeller-Plesset 2 A Møller-Plesset method truncated at third order. It costs substantially more than second order and does not reliably give a better answer, because the perturbation series does not converge smoothly. It is rarely used in practice. (UNVERIFIED) MP3 Moeller-Plesset 3 A Møller-Plesset method truncated at fourth order, including single, double, triple and quadruple excitations out of the Hartree-Fock reference. The triple excitations are what make it markedly more accurate than second order and also what make it expensive. It is sometimes written with the excitation levels listed explicitly. (UNVERIFIED) MP4 Moeller-Plesset 4 A Møller-Plesset method truncated at fifth order. It includes still higher excitations than fourth order, at a cost that rises steeply with system size. It is used almost only to study how the perturbation series itself behaves. (UNVERIFIED) MP5 Moeller-Plesset 5 A class of advanced quantum mechanics methods required for accurately describing molecules or chemical processes where the electronic structure cannot be well represented by a single dominant electron configuration (as assumed in basic Hartree-Fock theory). These situations often involve stretched bonds, excited states, diradicals, or transition metals, necessitating the use of multiple electronic configurations to capture the essential physics. These methods are typically more computationally demanding than single-reference approaches. (UNVERIFIED) multi-reference multi-reference method A multi-configurational self-consistent field method that includes every possible electron configuration within a chosen set of electrons and orbitals, the active space, and optimises the orbitals at the same time. The remaining orbitals stay doubly occupied or empty. Choosing the active space is the main difficulty, because the cost grows factorially with its size, and the method is the most widely used multi-reference starting point. (UNVERIFIED) CASSCF complete active space self-consistent field A quantum chemistry method that optimises the shapes of the molecular orbitals and the weights of several electron configurations at the same time. A single configuration cannot describe a bond being broken or a near-degenerate ground state; several can. It is the starting point for the multi-reference correlated methods. (UNVERIFIED) MC-SCF multi-configurational self-consistent field A quantum chemistry method that applies a coupled cluster treatment with single and double excitations on top of a multi-configurational reference wavefunction. It combines the size consistency of coupled cluster with a reference able to describe bond breaking. It is among the most accurate methods available and among the most expensive. (UNVERIFIED) MR-CCSD multi-reference coupled cluster with single and double excitations A quantum chemistry method that performs a configuration interaction treatment with single and double excitations starting from a multi-configurational reference rather than a single determinant. It recovers dynamic correlation on top of the static correlation the reference already describes. It is not size-consistent, which limits its use on large or dissociating systems. (UNVERIFIED) MR-CI(SD) multi-reference configuration interaction with single and double excitations A multi-configurational self-consistent field method that divides the active space into subspaces and limits how many electrons may be excited out of or into each, while optimising the orbitals at the same time. The restriction keeps the number of configurations manageable where a complete treatment would be too large. It trades some completeness for the ability to use a bigger active space. (UNVERIFIED) RASSCF restricted active space self-consistent field A post-Hartree-Fock method that adds terms quadratic in the excitation coefficients to the truncated configuration interaction equations, so that the energy becomes size-consistent. Size consistency means the energy of 2 well-separated molecules equals the sum of their separate energies, which truncated configuration interaction does not give. The method is closely related to coupled cluster theory. (UNVERIFIED) quadratic configuration interaction A quadratic configuration interaction method with single and double excitations that adds a perturbative estimate of triple excitations. The triples correction is what brings the method close to chemical accuracy for many properties. It costs considerably more than the calculation without it. (UNVERIFIED) QCISD(T) A post-Hartree-Fock method that includes single and double excitations in a configuration interaction treatment modified so that the result is size-consistent. Size consistency means the energy of 2 well-separated molecules equals the sum of their separate energies, which plain truncated configuration interaction does not satisfy. Its cost and accuracy are close to those of coupled cluster with single and double excitations. (UNVERIFIED) QCISD A computational method that uses a simplified energy function whose parameters were fitted to experimental measurements or to more accurate calculations, rather than derived from first principles. The fitting is what makes it fast enough for large systems. It is also its limit: the method is reliable only for systems resembling those it was fitted to. (UNVERIFIED) empirical method A method that adjusts the positions of atoms to reach a nearby arrangement of lowest energy. It moves downhill from wherever it starts, so it finds the nearest low point rather than the lowest one anywhere. It is run before a simulation to remove the severe atom overlaps that a freshly built system almost always contains. (UNVERIFIED) minimization method Molecular dynamics (MD) is a computer simulation technique used to study the physical movement of atoms and molecules over time. It calculates the forces between particles using an empirical potential energy function (force field) and then integrates Newton's equations of motion numerically to simulate how the positions and velocities evolve in small time steps. MD simulations provide detailed information about the dynamic behavior, conformational changes, and thermodynamics of molecular systems. MD molecular dynamics The NetCDF (Network Common Data Form) format, which typically uses the .nc extension, is a set of software libraries and a file format designed for storing and sharing array-oriented scientific data. Developed and maintained by the Unidata program at the University Corporation for Atmospheric Research (UCAR), it is an open standard widely used across many scientific disciplines, such as molecular dynamics. While Amber has adopted NetCDF as its preferred, modern format for storing trajectories and other data, the format itself is not specific to Amber and is a general-purpose standard. NetCDF format Enhanced sampling encompasses a diverse set of computational techniques used in molecular simulations to accelerate the exploration of a system's possible configurations or states. These methods aim to overcome the timescale limitations of standard simulations, particularly when high energy barriers separate important states or when rare events need to be sampled efficiently. They often involve modifying the potential energy surface, using multiple simulation copies, or applying biasing forces. enhanced sampling An enhanced sampling method that gradually builds up a force opposing the system's own along a chosen coordinate, until the system moves along that coordinate freely. Once the opposing force matches the true one, the barriers along that coordinate have been flattened. The accumulated force is itself the answer, because it gives the free energy profile. (UNVERIFIED) adaptive force bias Canonical adiabatic free energy sampling refers to enhanced sampling strategies where system parameters or constraints are changed very slowly (adiabatically) during a simulation performed under constant NVT conditions. It splits a system into a "subsystem" (the reactive part) and an "environment". By introducing fictitious masses and decoupling the subsystem, it allows the subsystem to evolve slowly and adiabatically on the potential of mean force while operating at a much higher temperature than the environment, speeding up rare events without altering the physical ensemble. canonical adiabatic free energy sampling Forward Flux Sampling (FFS) is an enhanced sampling technique specifically designed to calculate rate constants and characterize pathways for rare events, such as transitions between stable molecular states. It involves defining a series of non-overlapping interfaces along a reaction coordinate between the initial and final states. Trajectories are initiated from these interfaces and only propagated forward if they reach the next interface, efficiently focusing computational effort on successful transition paths. forward flux simulation Locally Enhanced Sampling (LES) is a computational method used in molecular dynamics to accelerate the simulation of specific, localized regions of a molecule, such as a ligand binding site or a specific protein segment. It achieves this by introducing multiple non-interacting copies of the "enhanced" atoms while proportionately reducing their interaction potentials. LES locally enhanced sampling A statistical model of a system that moves between discrete states, in which the next state depends only on the current one and not on how the system arrived there. That assumption is what makes long-timescale behaviour computable from short observations. It holds only if the states are chosen so that everything relevant about the past is captured by which state the system is in. (UNVERIFIED) markov chain model The max-flux approach in molecular systems typically refers to a variational method used to calculate the most probable reaction pathway during a transition (like conformational changes or chemical reactions). It finds the path that maximizes the diffusive flux of distinct reactive trajectories between two metastable states. max-flux approach Metadynamics is an enhanced sampling technique that accelerates the exploration of complex energy landscapes by adding a history-dependent bias potential along a few chosen collective variables (CVs). Small repulsive energy 'hills' (often Gaussian functions) are periodically deposited at the system's current location in the CV space, discouraging revisiting explored areas and eventually filling up free energy wells. The accumulated bias potential provides an estimate of the free energy surface along the selected CVs. metadynamics Milestoning sampling (or the milestoning algorithm) is an advanced molecular simulation technique designed to calculate the kinetics and thermodynamics of slow, complex processes (such as protein folding or drug binding). It breaks down computationally expensive simulations into multiple, short, independent trajectories between predefined interfaces, or "milestones". milestoning A free energy analysis method that combines data from simulations run at many thermodynamic states at once, and returns the free energy difference between any pair of them together with equilibrium averages. It uses every sample from every state rather than only neighbouring pairs, which is what makes it statistically optimal. It generalises the 2-state Bennett acceptance ratio. (UNVERIFIED) multi-state bennett acceptance ratio A machine learning model built from layers of simple units, each combining its inputs and passing the result on, whose connection strengths are adjusted to fit training data. Given enough units it can represent almost any relationship, which is why it is used where no formula is known. In this field it is used to predict energies and forces, and to find useful coordinates automatically. (UNVERIFIED) neural networks neural network The Nudged Elastic Band (NEB) method is primarily used to find the Minimum Energy Path (MEP) connecting two known stable states (e.g., reactant and product) on a potential energy surface. It works by optimizing a discrete chain of intermediate structures (images) linking the start and end points. Forces acting on the images perpendicular to the path are minimized, while spring forces along the path ensure continuity, guiding the chain towards the MEP, including the transition state. NEB nudged elastic band An enhanced sampling method that transforms the coordinates so that regions with known energy barriers, described by a simpler reference potential, take up less of the configuration space and low-energy basins take up more. The transformation leaves the partition function unchanged, so averages computed with the transformed coordinates need no reweighting afterwards. This makes those barriers easier to cross without changing the equilibrium properties of the system. (UNVERIFIED) reference potential spatial warping Replica-exchange molecular dynamics (REMD), also known as parallel tempering, is an enhanced sampling technique that runs multiple simulations (replicas) of the same system in parallel, typically each at a different temperature. At regular intervals, attempts are made to swap the current configurations between pairs of replicas running at adjacent temperatures, accepted based on a probability criterion. This allows configurations explored at high temperatures (overcoming energy barriers) to diffuse down to lower temperatures, enhancing conformational sampling. REMD replica-exchange molecular dynamics Steered Molecular Dynamics (sMD) is a simulation technique where an external, time-dependent force or constraint is applied to manipulate the system, often by pulling on specific atoms or controlling the distance between groups. This forces the system along a chosen pathway, allowing exploration of processes like ligand unbinding or protein monomerization-multimerization that might not occur spontaneously on simulation timescales. sMD trajectories can also be used with fluctuation theorems (like Jarzynski's equality) to estimate equilibrium free energy differences. steered dynamics A computational method that finds the most likely route between 2 stable states of a system, described in terms of a few chosen coordinates. It represents the route as a line of points and repeatedly moves them downhill on the energy or free energy surface while keeping them evenly spaced along it. The line settles onto the path the system is most likely to take, which is rarely a straight one. (UNVERIFIED) string method The swarm-of-trajectories approach refers to computational strategies that utilize a large number of independent, often relatively short, MD simulations launched from specific initial conditions. This "swarm" allows for efficient statistical sampling of pathways, transition probabilities between states, or average forces needed in methods like the String Method or Milestoning. It contrasts with relying on a single, very long trajectory to observe events of interest. swarm-of-trajectories Targeted Molecular Dynamics (TMD) is a simulation technique used to induce a conformational transition between a known initial structure and a known target structure. It achieves this by adding an extra potential energy term that biases the system to reduce its root-mean-square deviation (RMSD) relative to the target structure over the course of the simulation. TMD provides a plausible pathway for the transition, although not necessarily the minimum energy or most probable one. TMD targeted molecular dynamics Transition Path Sampling (TPS) is a powerful simulation technique designed to study rare transition events between stable molecular states by directly sampling the ensemble of reactive pathways connecting them. Starting with one known reactive trajectory, it uses a Monte Carlo sampling in path space (generating new trial paths from old ones) to explore the set of all possible transition mechanisms without needing predefined reaction coordinates. TPS allows for the calculation of rate constants and detailed mechanistic analysis. TPS transition path sampling The Transition State Ensemble (TSE) refers to the collection of molecular configurations that lie at the top of the highest free energy barrier separating the reactant and product states of a process, such as protein folding or a chemical reaction. Structures within the TSE are equally likely to proceed forward to the product state or revert back to the reactant state (i.e., they have a committor probability of 0.5). Characterizing the structural features of the TSE is key to understanding the mechanism of the transition. TSE transition state ensembles transition state ensemble Transition Interface Sampling (TIS) is a path sampling technique, related to Forward Flux Sampling, used for calculating rate constants of rare events. It involves defining interfaces along an order parameter and then specifically sampling ensembles of trajectories that start in the reactant state and reach a particular interface, or start at one interface and reach the next. By calculating the crossing probabilities between interfaces, TIS reconstructs the overall rate constant. TIS transition interface sampling An input file format that carries the keywords and data blocks instructing the Amsterdam Density Functional software how to perform a quantum chemical calculation. It specifies details like the desired computational task (e.g., geometry optimization, frequency calculation), the molecular geometry, the choice of density functional and basis set, and other calculation parameters. The format uses specific ADF keywords and block structures (e.g., ATOMS, GEOMETRY). (UNVERIFIED) ADF input format A CHARMM input file is a script containing a sequence of commands written in the CHARMM scripting language, which directs the CHARMM program to perform various tasks. These tasks can range from reading topology and coordinate files, setting up system parameters, running energy minimizations or molecular dynamics simulations, to analyzing results. CHARMM input format An input file format that instructs the GAMESS (US) quantum chemistry program, structured into groups identified by keywords starting with '$', such as $CONTRL, $BASIS and $DATA. Within these groups, keywords specify the type of calculation (e.g., RUNTYP=OPTIMIZE), the theoretical method (e.g., SCFTYP=RHF), basis set, molecular geometry, and other computational details. The format follows specific Fortran namelist-like conventions. (UNVERIFIED) GAMESS input format A text data format that specifies a molecular system and the settings for a quantum chemistry calculation to the GAMESS-UK program, using directives such as TITLE, RUNTYPE, GEOMETRY and BASIS. Its keywords and layout differ substantially from those of the separate GAMESS (US) program, so files are not interchangeable between them. (UNVERIFIED) GAMESS-UK input format An input file format that instructs the Gaussian quantum chemistry software, carrying the extension .com or .gjf. It characteristically includes a route section, starting with #, containing keywords that define the theoretical method, basis set and calculation type, followed by a title section, the charge and spin multiplicity, and the molecular geometry. (UNVERIFIED) GAUSSIAN com format GAUSSIAN gjf format gjf format In GROMACS, the primary molecular dynamics input file usually has an .mdp extension and contains keyword-value pairs defining the simulation parameters. Processed by the grompp tool, it controls aspects like the integration algorithm, timestep, simulation length, temperature and pressure coupling methods, and cutoff schemes. This file dictates how the MD simulation engine (mdrun) will execute the run. GROMACS input format A data format, written with the extension .hin, that is the native way the HyperChem modelling package stores a molecular system. It holds atomic coordinates, connectivity, force field atom types, and settings for a calculation. It serves both as input to the program and as the way a session is saved. (UNVERIFIED) HyperChem hin format An input file format that carries the instructions for a quantum chemistry calculation in Jaguar, part of the Schrodinger suite. It uses a specific syntax with data sections delimited by &section_name (e.g., &gen, &zmat) and terminated by &, containing keywords and values controlling the calculation type, method, basis set, and molecular geometry. The structure facilitates setting up various QM tasks. (UNVERIFIED) Jaguar input format A data format specifying a molecule's geometry to the MOPAC semi-empirical program as 3 coordinates per atom. Each coordinate carries a flag saying whether it may be optimised or must be held fixed. That per-coordinate control is what distinguishes it from a plain coordinate list. (UNVERIFIED) MOPAC cartesian format A data format specifying a molecule's geometry to the MOPAC semi-empirical program as bond lengths, angles and rotations. As with its coordinate counterpart, each value carries a flag saying whether it may be optimised. It is the form used when a particular bond or angle must be scanned or held. (UNVERIFIED) MOPAC internal format A NAMD configuration file, typically ending in .namd, contains parameters and commands that control a molecular dynamics simulation run using the NAMD software. It uses a keyword-value pair format, often leveraging Tcl scripting capabilities, to specify input files (structure, coordinates, parameters), simulation conditions (temperature, pressure, duration), and algorithms. This file dictates the entire simulation protocol. NAMD configuration format An NWChem input file (.nw) uses directives and nested blocks to instruct the NWChem software package on performing computational chemistry tasks. It is composed of commands, called directives, which define data (such as basis sets, geometries, and filenames) and the actions to be performed on that data. Directives are processed in the order presented in the input file, with the exception of certain start-up directives (see Input File Structure) which provide critical job control information, and are processed before all other input. Most directives are specific to a particular module and define data that is used by that module only. A few directives (see Top-level Directives) potentially affect all modules, for instance by specifying the total electric charge on the system. NWChem input format An input file format that instructs the Q-Chem quantum chemistry program, using data blocks marked by $block_name and $end. Key blocks include $molecule (defining charge, multiplicity, and geometry), $rem (specifying calculation method, basis set, job type via keywords), and potentially others like $basis or $opt. This structure organizes the input required for a wide range of QM calculations. (UNVERIFIED) Q-Chem input format A text data format in which the Amsterdam Density Functional program records a completed calculation. It holds the input as read back, the convergence history, the final energy with its breakdown into components, the final geometry and computed properties such as charges and dipole moments. (UNVERIFIED) ADF output format A binary data format, written by the Amsterdam Density Functional program with the extension .rkf and called TAPE21 in versions up to 2019, that stores the results of a calculation as named sections that can be read by keyword. It holds information such as the geometry, basis set details, molecular orbital coefficients and computed properties. It is read for analysis and visualization, and can be reused as a fragment file in a later calculation on a larger molecule. (UNVERIFIED) ADF Tape41 format ADF rkf format RKF format The CHARMM output file is the main text log generated during a CHARMM run, typically containing an echo of the input script commands interspersed with the results produced by those commands. This includes detailed energy breakdowns, progress information during dynamics or minimization (like step number, time, temperature, pressure, energies), and outputs from analysis routines. It serves as the primary record of the CHARMM job execution. CHARMM output format A text data format, written with the extension .log or .out, in which the GAMESS (US) program records a completed calculation. It holds the input as read back, the basis set and geometry, convergence progress for the SCF or geometry optimisation, final energies, molecular orbital information and calculated properties. It is the file results are extracted from. (UNVERIFIED) GAMESS output format A text data format in which the GAMESS-UK program records a completed calculation, including the input as read back, convergence progress, final energies and calculated properties. Its structure and keywords differ from those of the separate GAMESS (US) program, so a parser written for one will not read the other. (UNVERIFIED) GAMESS-UK output format A text data format, written with the extension .log, in which the Gaussian program records a completed calculation. It holds setup information, convergence progress, final energies, molecular orbital details, properties such as dipole moments and charges, and vibrational and thermochemical results when those were requested. A compact summary called the archive entry sits near the end. (UNVERIFIED) GAUSSIAN log format The GROMACS MD output typically refers to the primary text log file (.log) generated by the mdrun simulation engine. It contains summary information about the simulation setup, detailed performance data (like PME load balancing and timings), step-by-step progress updates (if configured in the .mdp file), and final run statistics including energy averages and runtime. This file is crucial for monitoring performance and verifying run completion. GROMACS output format A text data format, written with the extension .out, in which the Jaguar program records a completed calculation. It holds the input parameters, convergence information for the SCF and geometry optimisation cycles, final energies, properties such as charges and dipole moments, and vibrational frequencies when requested. The final geometry is usually written to a separate structure file. (UNVERIFIED) Jaguar output format An output file format holding the results of a calculation performed by the MOPAC semi-empirical program, written with the extension .out. It typically includes an echo of the input keywords, details of geometry optimization steps (if performed), the final optimized geometry, the calculated heat of formation (ΔHf), ionization potential, dipole moment, and atomic charges. A summary section usually highlights the key results and timings. (UNVERIFIED) MOPAC output format A text data format, written with the extension .out or .nwout, in which the NWChem program records the tasks its input file requested. It holds the calculation setup, convergence progress, energies from whichever quantum, molecular mechanics or combined method was used, properties and timings. Because the program accepts very varied inputs, the layout of the file varies with them. (UNVERIFIED) NWChem output format A text data format, written with the extension .out or .qcout, in which the Q-Chem program records a completed calculation. It holds the method and basis set used, convergence information for the SCF and geometry optimisation, final electronic energies, molecular orbital information, and properties such as charges and dipole moments. A summary section usually collects the main computed values. (UNVERIFIED) Q-Chem output format A binary checkpoint file format that stores the state of a completed or running Gaussian calculation, including the molecular geometry, basis set details, molecular orbital coefficients, density matrix and forces. It is essential for restarting interrupted calculations (especially geometry optimizations or frequency calculations) and for post-processing tasks like generating cube files or performing wavefunction analysis. It can be converted to a formatted version (.fchk). (UNVERIFIED) GAUSSIAN chk format A binary data format, written with the extension .chk, in which Gaussian stores the state of a calculation so that a later job can continue from it. It holds the geometry, the wavefunction and Hessian information that a later job can read back. (UNVERIFIED) GAUSSIAN restart format A binary data format, written with the extension .cpt, in which GROMACS stores the complete state of a running simulation at intervals. It holds the step number, the time, the positions, the velocities, the box and the state of the thermostat and barostat. Resuming from it continues the run exactly, as though it had never stopped. (UNVERIFIED) GROMACS restart format Restarting an NWChem calculation involves using various files generated during a previous run, depending on the calculation type. This may include the run database (.db), molecular orbital vector files (.movecs), geometry files, or specific QM/MM restart files (.rst). NWChem restart format A data format that stores the force on every atom of a system at each saved step of an AMBER simulation. It records forces where a trajectory format records positions. It is used for debugging a force field and for analyses that need the forces themselves. (UNVERIFIED) AMBER mdfrc format mdfrc format The GROMACS .gro file is a simple, formatted text file used to store molecular coordinates, atom/residue names, and optionally velocities. It includes a title line, the total number of atoms, and then one line per atom specifying residue and atom information, coordinates (in nanometers), and optionally velocities. The file concludes with the periodic box vectors. It is typically used to store single structures. GROMACS gro format gro format A text data format used by the TINKER software to store a molecular system, giving for each atom its index, name, x, y and z coordinates, force field atom type and the indices of the atoms bonded to it. Although it shares the .xyz extension with the plain XYZ format, it is not interchangeable with it. It therefore serves as both a coordinate file and a topology file. (UNVERIFIED) TINKER xyz format A data format in which TURBOMOLE records the Cartesian coordinates of a molecule, as a data group headed $coord that usually sits in a separate file named coord, which the control file refers to. Each line gives the x, y and z coordinates of 1 atom in atomic units (bohr), followed by its element symbol. (UNVERIFIED) TURBOMOLE coordinate format A GROMACS include topology file (.itp) contains the topology definition for a specific type of molecule (like a protein, lipid, or solvent). It typically includes sections like [ moleculetype ], [ atoms ], [ bonds ], etc., and is designed to be reusable. These .itp files are incorporated into the main system topology (.top) file using the #include directive. GROMACS include topology format GROMACS itp format itp format A topology file format that describes an entire molecular system to be simulated with GROMACS, written with the extension .top. It typically starts by including force field parameter files and specific molecule topology files (.itp) using #include directives. Crucially, it concludes with a [ molecules ] section listing the names and counts of each type of molecule present in the simulation box. (UNVERIFIED) .top GROMACS system topology format GROMACS top format top format A data format in which NWChem stores the topology of a system for molecular dynamics or QM/MM calculations, as a separate file that its prepare module generates from the input coordinates and the force field files. It holds the static description of the system: the atoms, the bonded interactions, the excluded pairs and the force field parameters. The quantities that change, such as coordinates and velocities, are kept in a separate restart file. (UNVERIFIED) NWChem topology format A Protein Structure File (.psf) defines the atoms, connectivity (bonds, angles, dihedrals, impropers), and basic atomic properties (like type indices, charges, masses) for a molecular system. Importantly, it does not contain the actual force field parameters. PSF files are used by the simulation programs CHARMM and NAMD. The format originates with X-Plor. Recorded here rather than in the definition, which was verified on 2026-09-18 without it. Source: EDAM 1.25, class format_3882 PSF (http://edamontology.org/format_3882). PSF format The DCD (.dcd) trajectory format is a widely used binary file format for storing atomic coordinates from molecular dynamics simulations, optionally including periodic box information. Originally developed for CHARMM, it is now supported by many simulation packages (like NAMD, LAMMPS) and analysis tools (like VMD) due to its efficiency and portability. dcd format A binary data format in which GROMACS records the positions of atoms through a simulation, and in some forms their velocities and forces as well. The compressed forms reduce the precision of the coordinates to save space, which is acceptable for analysis but not for restarting a run exactly. Which form was written determines what the trajectory can afterwards be used for. (UNVERIFIED) GROMACS trajectory format A molecular structure is the information that specifies the three-dimensional arrangement of atoms in a molecule. It is typically represented by a list of atomic coordinates (X, Y, Z) for each atom in the system. This structural information is the fundamental input for visualization and is the starting point for any molecular simulation. molecular structure OBSOLETE. Small molecule structure formats are designed primarily for representing organic and drug-like molecules, often emphasizing chemical details like bond orders, formal charges, and stereochemistry. These formats are widely used in cheminformatics, drug discovery databases, and as input for various modeling tasks. Common examples include SDF, Mol2, and SMILES strings. Retired in the file-format flattening. Classifying structure formats by molecule size was an artificial distinction: a small molecule can be stored in a PDB file and a macromolecule in a mol2 file. Its former children are now direct children of 'molecular structure format' (MOLSIM:001088). Reported in GitHub issue #20. obsolete small molecule structure true The Mol2 format (.mol2) is a text-based file format originating from Sybyl software, commonly used to represent small molecules, especially in docking and cheminformatics applications. It stores atomic coordinates, atom names, connectivity (including bond orders), partial charges, and often uses specific Sybyl atom types. A single Mol2 file can contain multiple molecular structures. Mol2 format The Structure-Data File (SDF, .sdf) format is widely used in cheminformatics for storing libraries of small molecules along with associated data. It essentially consists of multiple blocks defining atomic coordinates and connectivity for each molecule concatenated together, where each block can be followed by data fields like IDs or calculated properties. The format was developed by MDL Information Systems. Recorded here rather than in the definition, which was verified on 2026-09-18 without it. Source: EDAM 1.25, class format_3814 SDF (http://edamontology.org/format_3814). SDF format Sequence formats are standardized ways to represent the primary structure of biological macromolecules, such as the order of amino acids in a protein or nucleotides in DNA/RNA, using strings of characters. These formats usually include a header line providing an identifier and description, followed by the sequence itself using standard codes (e.g., one-letter codes). They are fundamental in bioinformatics for storing and exchanging sequence data. sequence data format A data format that stores a property varying continuously through 3-dimensional space, such as electron density or electrostatic potential, as values on a regular grid of points. A file in one of these formats states the grid's origin, its dimensions and its spacing, then lists a value for each point. They are used to visualize and analyse spatial distributions from simulation or from experiment. (UNVERIFIED) volumetric data format General data formats are versatile, often standardized ways to structure and exchange data that are not tied to a specific application domain like molecular simulation. Formats like XML or JSON use tags or key-value pairs to create hierarchical and self-describing data structures. They are sometimes employed in simulation contexts for storing metadata, configuration, or for facilitating data exchange between different tools. general data format A script holding the sequence of commands that an analysis program or library carries out on simulation data. It records exactly what was computed and in what order, so an analysis can be repeated. It is often the only surviving record of how a published number was obtained. (UNVERIFIED) analysis script The Fasta format is a simple and widely used text-based format for storing biological sequence data (nucleotides or amino acids). Each sequence entry begins with a single-line description header, marked by a greater-than symbol ('>'), followed by the sequence data on one or more subsequent lines. Its simplicity makes it easily readable by humans and parsable by many bioinformatics tools. FASTA format A text data format storing a property sampled on a regular 3-dimensional grid, together with the positions of the atoms it surrounds. It states the grid's origin, its axes and its spacing, then lists one value per point. Electron density, electrostatic potential and molecular orbitals are all distributed in it. (UNVERIFIED) GAUSSIAN cube format cube format XML (Extensible Markup Language) is a flexible, text-based markup language used to encode data in a format that is both human-readable and machine-readable. It uses user-defined tags to structure data hierarchically with elements and attributes. XML is widely used for data exchange on the web and in various scientific domains, including bioinformatics. XML format A data format, built on XML, that records chemical information as structured, labelled text. It defines tags for molecular structures, reactions, spectra and calculation results, so that a file states what each number is rather than relying on position. Its purpose is exchange between programs that would otherwise each need a reader for the others' formats. (UNVERIFIED) CML format A text data format holding a sequence of commands for the ptraj or cpptraj trajectory analysis programs of AmberTools. A file in this format names the input trajectories with trajin, may name a reference structure, and then lists analysis commands such as rms, distance, hbond, radgyr or cluster. It is the record of how a trajectory was post-processed. (UNVERIFIED) Ptraj script format A category of analysis methods focused on examining the three-dimensional arrangement of atoms (geometry and conformation) and the overall shape and topology of molecules based on simulation data or optimized structures. This includes measuring geometric parameters (distances, angles, dihedrals), identifying characteristic local and global structural features, comparing structures, and quantifying structural fluctuations. It helps understand molecular architecture, flexibility, stability, and structure-function relationships. structural analysis OBSOLETE. A structural analysis focusing on the overall architecture, shape, fold, and packing of a macromolecule or molecular complex, rather than fine details. This is often characterized by metrics like the radius of gyration, principal moments of inertia, overall dimensions, secondary structure content (for proteins), or tertiary arrangement of domains. It provides a macroscopic view of the molecule's conformation and large-scale structural changes. Retired because it grouped nothing: the analyses its definition named (radius of gyration, principal moments of inertia, secondary structure) are siblings under MOLSIM:000917 structural analysis, not subclasses of this term. Raised in the August 2026 definition review. obsolete global structure analysis true A generic name for a text-based file format that stores the Cartesian (x, y, z) coordinates of atoms in a molecular system. The specific layout can vary, as multiple simulation packages like CHARMM and Amber have their own distinct versions of a ".crd" file. Due to its ambiguity, it is often necessary to specify which software's CRD format is being used. crd format A structural analysis focusing on the detailed geometric arrangement of small groups of atoms within a molecule. This includes analyzing specific bond lengths, bond angles, dihedral angles, ring conformations (like sugar pucker), hydrogen bonding patterns, or the conformation of amino acid side chains. It provides fine-grained insights into specific interactions, conformational preferences, and flexibility at the atomic or residue level. local structure analysis A fundamental local structure analysis that calculates the angle formed by three specified atoms (A-B-C), centered on atom B. These bond angles are essential parameters defining the geometry and shape of molecules. Monitoring their values and distributions during simulations reveals information about molecular flexibility and conformational preferences around specific centers. (UNVERIFIED) atom triplet angle calculation analysis A local structure analysis technique that calculates the torsion angle defined by four sequentially bonded atoms (A-B-C-D), representing the rotation around the central bond (B-C). Dihedral angles are critical for determining the three-dimensional conformation of flexible molecules, including protein backbones, side chains, and polymer chains. Analyzing their values and distributions reveals accessible rotational states and conformational dynamics. dihedral calculation analysis A basic local structure analysis technique that measures the distance between any two specified atoms (bonded or non-bonded) in a molecular system. Monitoring distances is essential for tracking bond lengths, identifying close contacts, analyzing non-covalent interactions (like hydrogen bonds, salt bridges, van der Waals contacts), and assessing overall conformational changes or molecular dimensions. distance calculation analysis A specialized local structure analysis used to describe and quantify the non-planar conformation (puckering) of cyclic molecular fragments, particularly five-membered (e.g., furanose rings in DNA/RNA, proline) or six-membered rings (e.g., pyranose rings, cyclohexane). It employs parameters like pseudorotation angles and amplitudes to precisely define the ring's shape (e.g., envelope, twist, chair, boat conformations). This analysis is vital for understanding the structure and function of nucleic acids and carbohydrates. (UNVERIFIED) pucker analysis A type of structural analysis focusing on quantifying properties associated with individual atoms within the molecular structure, often reflecting their local environment or dynamics. Examples include atomic fluctuations (Root Mean Square Fluctuation, RMSF, or B-factors) indicating mobility, solvent-accessible surface area (SASA) quantifying exposure to solvent, or coordination numbers describing the local packing density. Analyzing these properties provides atom-specific insights into dynamics, interactions, and structural roles. (UNVERIFIED) atomic properties analysis A force field for DNA from the Olomouc group that combines OL15 and OL21 and adds a refinement of the torsion terms that control sugar puckering (delta and tau1). The sugar pucker affects the balance between the A and B forms of DNA. It is the DNA force field recommended in AMBER. (UNVERIFIED) AMBER OL24 An analysis method that groups the structures in a trajectory by how similar they are, usually measured by the distance between them after superposition. It reduces thousands of frames to a handful of representative shapes. Those representatives are what a trajectory is usually described and pictured by. (UNVERIFIED) structure clustering analysis A specialized structure clustering approach where conformations are grouped primarily based on the values of selected dihedral angles, rather than overall coordinate RMSD. Structures are assigned to bins based on which ranges their key dihedral angles fall into, and clusters are formed by grouping structures within the same or similar bins. This method is particularly useful for analyzing conformational changes driven by specific torsions, like backbone transitions or sidechain rotations. (UNVERIFIED) structure clustering with dihedral binning Dynamics analysis involves processing the time-series data generated from molecular dynamics simulations, primarily the trajectory files, to extract quantitative information and insights about the system's behavior. This encompasses a wide range of techniques aimed at understanding molecular motions, conformational changes, interactions, and kinetic or thermodynamic properties. The goal is to translate raw simulation output into meaningful physical or biological understanding. dynamics analysis A text-based file format used by the AMBER software suite to store standard force field parameters. These files (historically using the .dat extension) contain the equilibrium values and force constants for bonds, angles, dihedrals, and non-bonded interactions for specific atom types. They serve as the central parameter database loaded by system-building tools like LEaP. (UNVERIFIED) AMBER dat format AMBER force field parameter format A coordinate rotation method is a structure modification method that applies a mathematical rotation to the coordinates of a molecule or a group of atoms. This procedure changes the orientation of the selected atoms in three-dimensional space without altering their internal structure. coordinate rotation modeling A dynamics analysis method that measures how the fluctuations of 2 different properties relate to each other over time, including at a time delay. A positive value means they rise and fall together, a negative value means one rises as the other falls, and near zero means they are independent. The delay at which the relationship is strongest shows which motion leads and which follows. (UNVERIFIED) time cross correlation analysis A dynamics analysis method that measures how the displacement of each atom or residue from its average position relates to that of every other, over the course of a simulation. A strongly positive value means 2 sites move in the same direction, a negative value means they move in opposite directions. The result is usually drawn as a dynamic cross-correlation map, which shows which parts of a molecule move as one. (UNVERIFIED) is it synonymous as "atomic correlation analysis"? RMSD correlation analysis Motion analysis encompasses techniques used to characterize the types, directions, and magnitudes of movements occurring within a molecular system during a simulation. These methods aim to simplify the complex, high-dimensional atomic motions into understandable patterns or dominant modes. This helps identify functionally relevant conformational changes, flexibility, and collective dynamics. motion analysis A structural analysis method that superposes structures by matching their principal axes of inertia rather than by minimising the distance between corresponding atoms. It aligns on overall mass distribution and shape, so it is unaffected by local differences that would dominate a least-squares fit. It suits comparisons of rotational motion and of overall shape. (UNVERIFIED) principal axis alignment analysis A dynamics analysis method that maps the many-dimensional motion in a trajectory onto 1 or a few chosen directions. The chosen direction may be a simple geometric measure such as a distance or an angle, or a vector taken from principal component analysis. Projecting this way reduces a trajectory to a small number of curves that can be read directly. (UNVERIFIED) coordinate projection analysis A dynamics analysis method that decomposes the motion of a molecular system into a set of independent collective modes, each with a characteristic amplitude or frequency. Principal component analysis applied to a trajectory and normal mode analysis applied to a single structure are the 2 standard ways of obtaining them. The modes show the principal directions in which a molecule moves. (UNVERIFIED) eigenmode analysis A type of dynamics analysis focused on quantifying the random translational and rotational movement of molecules or particles within the simulation environment due to thermal energy. It primarily involves calculating diffusion coefficients, which measure the rate at which particles spread out over time. This analysis provides insights into transport properties, viscosity, and mobility in fluids, membranes, or crowded cellular environments. (UNVERIFIED) diffusion analysis An analysis method that extracts a diffusion coefficient from a trajectory, most often by tracking how the mean squared displacement of the molecules grows with time. Over long enough times that growth is linear, and its slope gives the coefficient. Reading a slope from too short a stretch of trajectory is the usual way this measurement goes wrong. (UNVERIFIED) diffusion calculation analysis A type of structural analysis focused on calculating and interpreting the potential energy associated with specific molecular structures or conformations obtained from simulations or modeling. This includes calculating single point energies, comparing energies of different conformers, or decomposing the total energy into contributing components. It helps assess the stability and likelihood of different structures. energy analysis An energy analysis that computes the net energy change when a molecular complex forms from its separated components in a stated reference state, such as the gas phase or solution. The result is the difference in potential energy or free energy between the bound complex and the unbound molecules, and a more negative value means stronger binding. Quantum mechanical methods and end-point methods such as MM-PBSA and MM-GBSA can estimate it, some of them including entropy and solvation to give a binding free energy. (UNVERIFIED) binding energy analysis An effective energy analysis that estimates the binding free energy of a molecular complex by combining molecular mechanics energies from simulation snapshots with a Poisson-Boltzmann calculation of the electrostatic solvation energy and a surface area term for the nonpolar solvation energy. The energies are averaged over many snapshots of the complex and of its separated parts. Solute entropy is not captured by this averaging and, when wanted, is estimated separately, for example by normal mode analysis. (UNVERIFIED) MM-PBSA analysis An effective energy analysis that estimates binding free energies by combining molecular mechanics energies with a generalized Born model for the electrostatic part of the solvation free energy and a surface area term for the nonpolar part. The generalized Born model approximates the Poisson-Boltzmann equation at lower cost, which makes the method suited to large numbers of calculations such as virtual screening. In other respects it is the same as the Poisson-Boltzmann variant, MM-PBSA. (UNVERIFIED) MM-GBSA analysis An energy analysis that computes the change in potential energy when 2 or more molecules or fragments are brought from infinite separation to their positions in a complex, with the geometry of the complex held fixed. It is usually the sum of the electrostatic and van der Waals interactions between the parts, or of their quantum mechanical counterparts such as electrostatics, exchange-repulsion, induction and dispersion. It normally leaves out the energy needed to distort each fragment on binding and the effect of solvation, so it is only 1 part of the total binding energy. (UNVERIFIED) interaction energy analysis A specific type of interaction energy analysis focused on quantifying the energetic contribution of individual or collective hydrogen bonds within or between molecules. This can be estimated using specialized energy decomposition schemes, analysis of electrostatic and van der Waals terms between donor-hydrogen-acceptor atoms from a force field, or derived from quantum mechanical calculations. Quantifying H-bond energies helps understand their role in stabilizing structures and mediating molecular recognition. (UNVERIFIED) hydrogen bond interaction energies analysis A specific interaction energy analysis quantifying the stabilizing energetic contribution from the non-covalent interactions (primarily van der Waals/dispersions and electrostatics) between adjacent nucleobases (DNA, RNA). These interactions occur between the aromatic base rings and are essential for the stability, geometry, and mechanical properties of nucleic acid structures. Stacking interaction energies are commonly estimated using quantum mechanical methods or specially parameterized molecular mechanics force fields. base pair stacking energies analysis This refers to one of the file formats used by Amber to store atomic coordinates. Its extension is commonly .inpcrd, .crd or .mdcrd. It contains a list of coordinates in ASCII, usually read alongside a topology file, and may optionally include periodic box information. AMBER inpcrd format AMBER prmcrd format Amber's standard ASCII coordinate trajectory file format. The "Coordinates Data Set" or "COORDS" in Amber is not a file format, but rather a specialized, in-memory data structure used by the cpptraj analysis program. It is designed to efficiently hold and manipulate sets of coordinate frames (i.e., structures) loaded from trajectory files during an analysis session. Since the COORDS data set is an internal, in-memory representation, it does not have a file extension. Users can interact with it using file-based commands: Loading: users populate a COORDS data set by loading data from standard Amber file formats, such as restart files (.rst7), PDB files, or trajectory files (.nc, .mdcrd), using the loadcrd command. Saving: users write the contents of a COORDS data set to a file using the crdout command. The output will be in a standard coordinate or trajectory format (e.g., Amber restart, PDB, etc.), not a special "COORDS format." AMBER crd format A category of analysis methods focused specifically on characterizing the structure, dynamics, and properties of the solvent molecules (most commonly water) surrounding a solute molecule within a simulation. This includes examining solvent density and distribution around the solute, analyzing hydrogen bonding networks within the solvent and between solvent and solute, calculating residence times of solvent molecules near the solute, and assessing the overall impact of the solvent on solute behavior. It provides crucial insights into solvation phenomena. (UNVERIFIED) solvent analysis A hydration analysis that divides the water around a solute into concentric layers, or shells, usually by distance from the solute atoms or surface, and characterizes the water in each shell. It reports, for example, the number of water molecules, their density, preferred orientations, hydrogen bonding patterns and residence times compared with bulk water. This gives a structured description of the local solvent environment. (UNVERIFIED) water shell analysis A distribution analysis that calculates the probability of finding a particle at a distance r from a reference particle or point, relative to the probability for a uniform distribution at the same density. The resulting radial distribution function, g(r), shows peaks for ordered solvation shells and approaches 1 at large distances, where the solvent behaves as bulk. Integrating it up to its first minimum gives the average number of neighbors in the first solvation shell. (UNVERIFIED) radial distribution analysis A specific type of validation technique focused on assessing the quality of the 3-dimensional arrangement of atoms (the conformation) in simulated biomolecules. It evaluates whether the molecular geometry, including bond lengths, angles, and steric arrangements, is plausible according to known chemical and physical rules. This ensures the structural models generated or sampled during a simulation are stereochemically sound. (UNVERIFIED) conformation validation analysis A structural analysis that plots the backbone dihedral angles phi and psi of each amino acid residue in a protein, to show whether they fall in regions that are sterically allowed or disallowed. It is used to check the quality and realism of a protein backbone structure. In simulations it also shows which backbone conformations, such as helix or sheet, a protein samples. (UNVERIFIED) ramachandran analysis A form of conformation validation that involves examining various geometric features of a biomolecular structure beyond just backbone torsion angles. This may include verifying appropriate bond lengths, bond angles, checking for unrealistic close contacts between atoms (clashes), or assessing other stereochemical parameters against standard values. It serves as a general quality control measure for the overall molecular model. structural check analysis Refers to the regular, local folding patterns occurring within the backbone of polymers like proteins or nucleic acids, such as alpha-helices and beta-sheets in proteins. These structures are defined by specific patterns of hydrogen bonds between backbone atoms and represent key building blocks of the molecule's overall three-dimensional architecture. While not a technique itself, analyzing secondary structure content and stability is often part of simulation validation and analysis. secondary structure assignment analysis Refers to the dynamic transformations or fluctuations observed in the secondary structure elements of a biomolecule during a simulation or over time. Monitoring these changes, such as the formation or dissolution of helices or sheets, provides insights into molecular flexibility, folding pathways, or responses to interactions. Analyzing these dynamics helps understand how molecular function relates to structural transitions. secondary structure changes analysis A specific type of solvent analysis focusing on the direct interaction of water molecules with a solute molecule. It involves characterizing the arrangement, orientation, dynamics, and energetics of water molecules in the immediate vicinity of the solute, often referred to as the hydration shell(s). Understanding hydration patterns is critical as these interfacial water molecules strongly influence the solute's structure, stability, dynamics, and interactions with other molecules. (UNVERIFIED) hydration analysis A solvent analysis technique focused on determining the average spatial arrangement or probability density of solvent molecules relative to the solute molecule, averaged over a simulation trajectory. This often involves calculating various distribution functions (e.g., radial, spatial, or angular) that quantify how the likelihood of finding a solvent molecule varies with position and/or orientation around the solute. It reveals preferred solvation sites and the overall structure of the solvent environment. (UNVERIFIED) distribution analysis A core mathematical technique from linear algebra used to simplify a square matrix by finding its characteristic values (eigenvalues) and directions (eigenvectors), effectively transforming it into a diagonal matrix. In biomolecular simulations, this is heavily used in methods like Principal Component Analysis (PCA) to identify dominant modes of collective atomic motion from covariance matrices, or in Normal Mode Analysis (NMA) to find vibrational frequencies from Hessian matrices. Diagonalization reveals the fundamental components or modes underlying complex data or system dynamics. (UNVERIFIED) matrix diagonalization analysis A coarse-grained model in which groups of atoms, or whole residues, are represented as beads joined by simple springs. The springs are chosen to reproduce the overall shape and flexibility of the assembly rather than any chemistry, so the model can cover structures far too large to simulate atom by atom. What it can show is how the parts move relative to each other; what it cannot show is anything that depends on a specific chemical group. (UNVERIFIED) coarse-grained elastic mesoscopic model A coarse-grained model that treats a polymer as a bendable filament described by 2 numbers: its total length, and the distance over which it stays pointing the same way. In a simulation it is built as a chain of beads whose bond and angle terms reproduce those 2 numbers. It is how DNA is modelled when what matters is how it coils and loops rather than which base is where. (UNVERIFIED) coarse-grained worm-like chain Experimental or computational methods used alongside Molecular Dynamics (MD) simulations to provide synergistic information, validate simulation accuracy, or compute properties challenging for standard MD alone. These techniques bridge simulation results with real-world measurements or provide deeper thermodynamic or structural insights. They help build a more comprehensive understanding of the molecular system under study. complementary techniques for MD An experimental Nuclear Magnetic Resonance (NMR) phenomenon that arises from the through-space dipolar coupling between nearby nuclear spins (commonly protons), used as a complementary technique to MD. Detecting NOEs allows measurement of short inter-proton distances (typically < 6 Å), providing crucial experimental restraints for determining or validating biomolecular structures and conformational ensembles in solution. MD simulations can be used to predict NOEs for comparison with experiments or refined using experimental NOE data. NOE nuclear overhauser effect (NOE) An experimental technique, either using X-rays (SAXS) or neutrons (SANS), where radiation is scattered by molecules in solution at low angles, providing information complementary to MD simulations. The resulting scattering profile is sensitive to the overall size, shape, and quaternary structure (oligomerization) of the molecules, averaged over the entire ensemble present in the sample. SAS data provides valuable low-resolution structural information that can be used to validate conformational ensembles generated by MD or to guide modeling of large-scale structures. (UNVERIFIED) SAS small-angle scattering Spectroscopy is the study of how matter interacts with energy, particularly electromagnetic radiation like light. By analyzing how energy is absorbed, emitted, or scattered by a substance, we can learn about its atomic and molecular structure, composition, and dynamics. It serves as a fundamental set of tools for observing the properties of molecules. spectroscopy A spectroscopic technique that probes how the bonds in a molecule stretch, bend and twist, by measuring absorption in the infrared or scattering of visible light. Each vibration has a frequency set by the masses involved and the stiffness of the bond, so the spectrum reports on chemical structure. Infrared absorption and Raman scattering are its 2 main forms. (UNVERIFIED) vibrational spectroscopy A spectrum recording how strongly a sample absorbs infrared light across a range of frequencies. Absorption occurs where the frequency matches a molecular vibration that changes the molecule's dipole, so the positions of the bands identify chemical groups. Its band positions can be compared with vibrational frequencies computed by quantum chemical or force field methods. (UNVERIFIED) infrared spectra A spectrum recording the small shift in frequency of light scattered by a sample, which reveals the vibrations that took or gave energy to the photons. It detects vibrations that change the molecule's polarizability, while infrared absorption detects those that change its dipole, so the 2 methods often show different bands. They are therefore complementary rather than alternatives. (UNVERIFIED) raman spectra A spectrum recorded with the illuminating light tuned close to, but not at, an electronic absorption band of the sample. Approaching an absorption greatly enhances the scattering from vibrations connected with it, so a chosen part of a large molecule dominates the spectrum. It is used to study a selected site, such as the active site of a complex biomolecule. (UNVERIFIED) pre-resonance raman spectra A spectroscopic technique measuring the difference in infrared absorption between left and right circularly polarized light. Only molecules that are not superimposable on their mirror images show such a difference, and the sign of each band reveals which form is present. It reports on 3-dimensional structure in solution, where ordinary infrared absorption reports only on which groups are there. (UNVERIFIED) vibrational circular dichroism A spectroscopic technique that probes transitions of electrons between energy levels, usually driven by ultraviolet or visible light. The energy absorbed or emitted reports on the spacing of those levels, which depends on bonding and conjugation. It is how colour, fluorescence and photochemical behaviour are studied. (UNVERIFIED) electronic spectroscopy A spectrum recording how strongly a sample absorbs light across the ultraviolet and visible range, as a function of wavelength. Absorption occurs where the photon energy matches a gap between electronic levels, so the positions of the bands report on bonding and conjugation. It is the measurement a computed electronic transition is compared against. (UNVERIFIED) UV/Vis UV/visible spectra A property giving the full profile of an absorption or emission band in an electronic spectrum, including its width and any structure within it, rather than only where its peak lies. The width reports how much the molecule's geometry changes when it absorbs light, and the structure within it reveals individual vibrations. A calculation matching only the peak position can still be wrong about both. (UNVERIFIED) electronic transition band shape A spectroscopic technique that probes the magnetic behaviour of nuclei or unpaired electrons, by observing how they respond in an applied magnetic field. The response depends on the local environment of each spin, which is what makes it report on structure. Nuclear magnetic resonance and electron paramagnetic resonance belong here. (UNVERIFIED) magnetic spectroscopy An interaction between nearby unpaired electron spins that favours their pointing in opposite directions. The paired arrangement has lower energy than the aligned one, so the system has little or no net magnetism. It occurs between metal centres in many metalloproteins and determines their magnetic signature. (UNVERIFIED) antiferromagnetic coupling A 3-by-3 array describing how strongly an electron spin interacts with a nearby nuclear spin, and how that strength depends on direction. Its directional part reports where the unpaired electron sits relative to the nucleus. It is what electron paramagnetic resonance spectra are interpreted through. (UNVERIFIED) hyperfine spectra tensors hyperfine coupling tensor A spectroscopic technique that probes transitions between a molecule's rotational energy levels, almost always in the gas phase where molecules turn freely. The spacing of the levels depends on how the molecule's mass is distributed, so the spectrum gives bond lengths and angles very precisely. It is among the most accurate structural measurements available and works only for small molecules. (UNVERIFIED) rotational spectroscopy An interaction between a molecule's vibration and its rotation, arising because a vibrating molecule changes shape and so changes how easily it turns. It means the 2 motions cannot be treated as independent. It shows in a spectrum as a shift in rotational line spacing that depends on which vibrational state the molecule is in. (UNVERIFIED) vibration-rotation coupling A spectroscopic technique that measures how a sample responds differently to left and right circularly polarized light. Only molecules that are not superimposable on their mirror images give such a difference, so these techniques report on handedness. Optical rotation, circular dichroism and Raman optical activity belong here. (UNVERIFIED) specialized techniques chiroptical spectroscopy A chiroptical spectroscopy technique that measures the small difference in the intensity of Raman scattering from chiral molecules in right versus left circularly polarized light. It combines the vibrational information of Raman spectroscopy with sensitivity to a molecule's handedness and 3-dimensional structure. It is especially useful for studying the shape of biomolecules in water, close to their natural conditions. (UNVERIFIED) raman optical activity The turning of the plane of polarized light as it passes through a sample of molecules that are not superimposable on their mirror images. The direction and size of the turn depend on which mirror form is present and in what excess. It is how the handedness of a sample is measured, and computing it correctly is a demanding test of a quantum method. (UNVERIFIED) optical rotations optical rotation A statistical analysis that smooths short-term fluctuations in a time series and shows longer-term trends by averaging a property over a time window that slides along the series. The average is recomputed each time the window moves forward. Watching the running average of a property such as total energy or temperature is a common way to judge whether a simulation has reached a stable equilibrium. (UNVERIFIED) running average calculation analysis A text-based file format used by the GROMOS simulation software to store the complete configuration of a molecular system. It is organized into data blocks (e.g., POSITION, VELOCITY, BOX) and contains atomic coordinates, periodic box dimensions, and optionally velocities. This format typically uses the .cnf file extension and is the primary input for coordinates in GROMOS simulations. (UNVERIFIED) cnf format g96 format GROMOS configuration format A file format for the MPDyn simulation program that stores the atomic coordinates of a molecular system. This format contains the x, y, and z positions for each atom, structured for use with the MPDyn software package. It provides the essential structural input for running simulations with MPDyn. (UNVERIFIED) MPDyn crd format The .trj format is an ASCII (text-based) trajectory file format used by the Amber molecular dynamics software suite and other analysis programs. Its fundamental purpose is to store a time-series of atomic coordinates, capturing the dynamic motion of a system during a simulation. The file contains sequential "frames" where each frame is a complete set of X, Y, and Z coordinates for every atom in the system at a specific point in time. It can also store periodic box dimensions, although this capability is sometimes limited. AMBER trj format The .trj format belongs to a family of older Amber text-based trajectory formats. Its role is best understood in comparison to its relatives: vs. .mdcrd: The .trj and .mdcrd formats are functionally identical. Both are safe, relatively unambiguous choices for storing Amber ASCII trajectories. A file with one extension can often be renamed to the other without causing issues in analysis software. vs. .crd: The .crd extension is highly ambiguous and problematic. It is the native format for the CHARMM simulation package and is interpreted as such by default in many tools. Using .trj avoids this critical format conflict. vs. .nc (NetCDF): The NetCDF format is the modern, binary, and recommended format for Amber trajectories. Compared to .trj, NetCDF (.nc) files are smaller, faster to read/write, and store data with higher precision. The .trj format is considered a legacy option. trj format A distance matrix in molecular simulation is a 2D data set that stores the pairwise distances between a selection of particles, such as atoms or residues, calculated from a simulation trajectory. It is a common form of analysis used to study conformational changes, structural stability, and the relative movement of different parts of a molecule. Common examples include: - General Text Formats: Often saved as plain text (.txt), comma-separated values (.csv), or a generic data (.dat) file for easy parsing and use in other programs. - Program-Specific Formats: Some simulation packages output the matrix in their own specific formats. For example, GROMACS can generate a distance matrix as an .xpm file, which is a portable pixmap image format. - Binary Formats: For efficiency, especially with large matrices, they may be saved in binary formats like NumPy's .npy format, which is common in Python-based analysis workflows. distance matrix format A text data format that stores a 2-dimensional matrix, such as a distance matrix, an RMSD matrix or a hydrogen bond pattern, by encoding each value as a character standing for a colour. It is a specialised use of the general XPixMap image format, so an ordinary image viewer can open it. GROMACS writes it and converts it to PostScript with gmx xpm2ps. (UNVERIFIED) .xpm GROMACS xpm format XPM format A text data format that holds the complete static description of a molecular system for an AMBER simulation: its atoms, the bonds between them, the force field parameters that govern their interactions, and their partial charges. It carries everything the simulation engine needs except the coordinates, which come from a separate file. It is read once at the start of a run and does not change during it. (UNVERIFIED) AMBER PARM7 format AMBER parm7 format AMBER parmtop format AMBER prmtop format parm7 format prmtop format OBSOLETE. A file format from the Amber simulation software that describes the structure and force field parameters of a molecule. This is the traditional, ASCII-based version of the Amber parameter/topology file, containing all information needed to calculate forces on the system. It is often used interchangeably with the prmtop format. AMBER PARM7 format AMBER parm7 format Retired 2026-08-07 as an internal duplicate of MOLSIM:001042 prmtop format. ParmEd writes one Amber topology format under two extensions, .prmtop and .parm7, so these were never two formats. Both terms also carried skos:exactMatch to the same EDAM term, EDAM:format_3881 AMBER top, which is what surfaced the duplication. prmtop survives because it is the name that actually occurs in practice: 19 occurrences against 0 for parm7 in the AMBER keyword corpus this ontology was extracted from. Its names are kept as exact synonyms on MOLSIM:001042. obsolete parm7 format true The Crystallographic Information File (cif) format is a standard text-based file format used to store information about molecular structures, especially those determined by experiments. It uses a flexible tag-based structure with key-value pairs, making it more robust and extensible for handling the large, complex structures and extensive metadata common in modern structural biology. As the official format of the Protein Data Bank (PDB), the macromolecular CIF (mmCIF) variant has superseded the older PDB format for archiving and distributing 3D coordinates of proteins and nucleic acids. CIF format A file format used by the TINKER simulation package to store the trajectory of a molecular system. It archives the atomic coordinates from different snapshots of a simulation in a single file for later analysis. This is the standard trajectory output for TINKER simulations. (UNVERIFIED) .arc TINKER arc format This can also be a coordinate format from Tinker package as well. arc format The original residue number is the number assigned to an amino acid in the original experimental data file. As a type of identifier, it represents the residue number as it appeared in the source file, typically a PDB file, before any renumbering may have occurred during simulation setup. This is crucial for mapping simulation results back to the standard numbering used in publications and other databases, especially when multiple chains or insertions are present. oresnums The residue number as it appeared in the original input (e.g., PDB) file original residue number A chain ID is a letter used to label different polymer molecules, like different protein chains, in a structure file. The chain ID usually has a limited number of characters like in the PDB format (1) or in mmCIF and PSF formats (4). It is an essential part of the atom selection syntax in analysis tools, allowing for targeting of specific molecules in a system. chainid An identifier used in PDB files to distinguish different polymer chains. chain ID An insertion code is a letter used in PDB and mmCIF structure files when extra amino acids have been inserted into a sequence without renumbering every residue after the insertion, thus allowing multiple residues to share the same residue number. This code is critical for unambiguously identifying specific residues in atom selections (e.g. residue "10A" is distinct from "10B"). This is referred to in PDB files as "icode". icodes A character used in PDB files to handle insertions in a sequence. insertion code A file format from the Amber software that saves the state of a simulation in plain text, allowing the run to be continued later. It contains the atomic coordinates, velocities, and periodic box dimensions at a specific point in time, in a human-readable form. The same ASCII format is written under several extensions, including .rst, .rst7 and .restrt. (UNVERIFIED) restrt format rst7 format rst format OBSOLETE. A text-based file format from the Amber software that saves the state of a simulation. It is the ASCII version of the Amber restart file, containing coordinates, velocities, and box information in a human-readable format. This format is often used for its portability and ease of inspection. Retired 2026-08-07 as an internal duplicate of MOLSIM:001049 rst format. ParmEd writes one Amber ASCII coordinate format under several extensions, .rst7, .restrt and .inpcrd, so .rst and .rst7 were never two formats; the real division in this family is ASCII against the binary NetCDF MOLSIM:001051 ncrst format. Both terms also carried skos:closeMatch to the same EDAM term, EDAM:format_3886 RST, which is the pattern that surfaced the duplication. The curator chose rst as the surviving label; note that ParmEd itself does not write a .rst extension, so rst7 format and restrt format are kept as exact synonyms on MOLSIM:001049. obsolete rst7 format true A binary data format from AMBER, built on the NetCDF container, that saves the state of a simulation at one moment so that the run can be continued. It stores the coordinates, the velocities, the box dimensions and the time with high precision. It is the default and recommended restart format in current AMBER versions. (UNVERIFIED) A NetCDF-based restart file for Amber simulations. ncrst format CHARMM commonly uses coordinate files in its own CARD format (.crd), which is a formatted ASCII file including atom coordinates along with atom numbers, residue names/numbers, atom types, and possibly weighting factors. It often starts with a title section describing the system. CHARMM crd format A file format from the CHARMM simulation software that stores the atomic coordinates of a molecule. It is a binary, unformatted version of the CHARMM coordinate file, making it more compact and faster to read than its text-based counterpart. This format is often used for saving and reading structures within CHARMM workflows. (UNVERIFIED) CHARMM cor format A binary file format from the GROMACS software that stores the full-precision trajectory of a simulation. It saves the atomic coordinates, velocities, and forces at each time step without any loss of precision. This format is used instead of XTC format when high accuracy is needed for analysis. GROMACS trr format trr trajectory format A compressed, binary file format from the GROMACS software for storing the trajectory of a molecular simulation. It uses a lossy compression algorithm to store atomic coordinates with reduced precision, resulting in much smaller file sizes. This is the most common trajectory format for GROMACS users when only atomic positions are needed for analysis. (UNVERIFIED) GROMACS xtc format XTC trajectory format A modern, binary file format from the GROMACS software used to store the results of a molecular simulation. It supports various compression algorithms and can store not only coordinates but also velocities, forces, and other trajectory data. This format is designed for high efficiency and flexibility in storing large simulation datasets. (UNVERIFIED) GROMACS tng format TNG trajectory format An input file format for sqm, the semi-empirical quantum chemistry program in AmberTools, holding a title line, a namelist of calculation settings, and then 1 line per atom giving its atomic number, atom name and Cartesian coordinates. Its main use is in deriving AM1-BCC partial charges for new molecules, a step that antechamber carries out by running sqm. (UNVERIFIED) .sqm An input file format for the semi-empirical QM program SQM. SQM input format A binary file format used by the Desmond simulation software to store the trajectory of a molecular system. It efficiently stores atomic coordinates, velocities, and other data in a block-based structure. This is the native trajectory output format for simulations run with Desmond. (UNVERIFIED) .dtr Desmond dtr format A trajectory file format used by the Desmond simulation software. dtr format A data format holding the lowest-energy conformations found by a low-mode (LMOD) conformational search in AmberTools, stored in order of energy with the global minimum first. It uses the standard formatted AMBER trajectory layout and is updated at intervals as the search proceeds. How many conformations it keeps, and within what energy window of the minimum, are set by the user. (UNVERIFIED) .conflib LMOD conflib format conflib format A binary data format built on HDF5 in which the MDTraj library stores trajectories together with the topology they belong to. Keeping both in one file removes the usual need to supply a separate topology when reading a trajectory. It is the native format of that library rather than a community standard. (UNVERIFIED) .h5 An HDF5-based trajectory file format used by the MDTraj library. MDTraj h5 format A data format for molecular simulation data, defined as a community standard on top of the general-purpose HDF5 container. It specifies where positions, velocities, forces and metadata are placed, so any program following it can read another's output. It belongs to no single simulation program, which is what distinguishes it from the engine-specific formats. (UNVERIFIED) .h5md H5MD format The PQR format is a modification of the standard Protein Data Bank (PDB) format used in computational biology to store atomic coordinate data along with atomic partial charges (Q) and radii (R). The name itself is a mnemonic, representing PDB + Q (charge) + R (radius). Its primary advantage is that it bundles the necessary structural and electrostatic parameters into a single, PDB-like file that is easily parsed by both analysis software and molecular visualization programs. These files are most commonly generated by the PDB2PQR software, which processes a standard PDB file, adds missing hydrogen atoms, and assigns atomic charges and radii based on a chosen force field. .pqr PQR format The Molecular Operating Environment (MOE) is a comprehensive, commercial software system for drug discovery, molecular modeling, and computational chemistry developed by the Chemical Computing Group. It integrates visualization with a vast array of tools for structure preparation, docking, pharmacophore modeling, and molecular dynamics setup within a single unified interface. Molecular Operating Environment MOE A flexible version of the 3-site TIP3P water model, in which the O-H bonds and the H-O-H angle are described by harmonic terms rather than held rigid. Flexibility is needed when vibrational spectra are computed, or when the model is used with a method that cannot handle constraints. It requires a shorter timestep, because bond vibrations are the fastest motion in the system. (UNVERIFIED) A flexible three-site water model. TIP3PFW A water model with 3 sites, derived from the SPC geometry, whose bonds and angle can stretch and bend. Its bonded parameters were refitted alongside the flexibility so that the liquid's density and diffusion stay correct. Making a rigid model flexible without refitting changes both. (UNVERIFIED) A flexible three-site water model. SPCFW A text data format in which the secstruct command of cpptraj writes secondary structure results as 1 line per frame, giving the frame number and then 1 character for each residue, in the layout the older ptraj program used. It fixes no file extension, and it is kept so that the output can still be read by tools written for ptraj. (UNVERIFIED) CPPTRAJ ptraj format ptraj format Xmgrace is a computer program used to make scientific graphs and plots from data. As a type of visualization tool, Xmgrace is a 2D plotting program for X Window Systems. Some simulation analysis tools, like GROMACS and cpptraj, can directly generate output files in a compatible format for easy plotting of time-series data like RMSD or energy. This tool is associated with file formats like .agr and .xvg. xmgrace A visualization tool that draws 2-dimensional and 3-dimensional graphs from data files, driven from the command line or from a script. Because it is scriptable it can be called from an analysis pipeline to produce figures without a person present. It is one of the standard ways simulation output is plotted. (UNVERIFIED) gnuplot A visualization program for displaying and animating molecular structures and trajectories in 3 dimensions. It handles systems of millions of atoms and includes Tcl and Python scripting, so analyses can be run on what is being shown. It is widely used to inspect molecular dynamics trajectories. (UNVERIFIED) VMD visual molecular dynamics A visualization plugin for VMD that draws and animates collective modes of motion from normal mode analysis or principal component analysis. It shows a mode as arrows on the structure, or as a trajectory moving along it. It turns an eigenvector into something that can be looked at. (UNVERIFIED) A VMD plugin for visualizing and analyzing normal modes. NMWiz A data format that records a 3-dimensional histogram as though it were the path of a single particle, writing the 3 histogram dimensions as the x, y and z coordinates of 1 atom over successive frames. That trick lets a volumetric distribution, such as where water sits around a solute, be opened in a molecular viewer as a point cloud. It is produced by the hist command of cpptraj. (UNVERIFIED) CPPTRAJ traj3d format The traj3d "pseudo-trajectory" is not a standard, independent file format but rather a specialized output option available within the hist command in Amber's cpptraj analysis program. It is created by taking the three data dimensions of a 3D histogram and writing them out as the X, Y, and Z coordinates of a single "atom" over time. This reformatting creates a trajectory-like file that can be loaded into molecular visualization software. The purpose is to allow researchers to visually inspect the 3D distribution of data from the histogram in a more intuitive, interactive 3D space, rather than as a static 2D plot. There is no mandatory file extension for the output file itself. traj3d format OBSOLETE. An information content entity is an abstract concept representing any piece of data, information, or knowledge that can be identified and described. In the context of simulations, it refers to the abstract data itself, such as the concept of a molecular structure or a simulation parameter set, separate from the specific file format used to store it. This distinction helps in organizing and standardizing the description of simulation data and metadata. Retired 2026-08-12 when MOLSIM adopted COB as its upper level. IAO:0000030 information content entity is one of COB's five roots and says the same thing, so keeping a MOLSIM class beside it was a duplicate. Its 8 direct children were re-parented onto IAO:0000030 in the same change; the 46 terms below them moved with their parents. obsolete information content entity true A molecular system specification defines the physical and chemical properties of a molecule required for computational chemistry and quantum mechanical simulations. This specification is the complete informational blueprint of the system before any simulation is performed. molecular system specification A process specification is the abstract set of instructions and parameters that define how a computational task, like a simulation or an analysis, should be performed. It represents the complete plan for the process, including the choice of algorithms, control settings, and input files. This specification must be recorded to ensure a computational experiment is reproducible. process specification An output file is generated by a computational program and contains the results of the calculation or simulation, along with logs, progress information, and any error messages. These files are crucial for analyzing the outcome of computational jobs. process output A molecular system specification that describes what a molecular system is made of: its atoms, their types and charges, the bonds between them, and the force field parameters that apply. In most simulations it does not change during the run, which is why it is kept separate from the coordinates. Without it, a set of coordinates cannot be read as molecules. (UNVERIFIED) molecular topology A category of data formats used by the CHARMM simulation package to describe the molecular topology of a system. These formats define the atoms, connectivity, and the chemical nature of residues. (UNVERIFIED) CHARMM topology format A data format used by CHARMM that defines a reusable template for one chemical building block, such as an amino acid or a nucleotide. It gives the atoms in that block, their types and charges, the bonds between them, and their standard geometry. A topology for a whole molecule is assembled by joining these templates in sequence. (UNVERIFIED) residue topology file (RTF) format rtf format A text data format used by AMBER to supply force field parameters that are missing from the main parameter files, or to override ones that are present. It lets a user add parameters for a non-standard molecule without editing the distributed force field. It is read in addition to the main files rather than in place of them. (UNVERIFIED) AMBER frcmod format frcmod format The complete collection of settings that tells a program how to run one simulation, including the temperature and pressure to hold, how long to run, how often to save, and which physical approximations to use. It is normally kept as a single input file. Together with the starting structure it is what would be needed to repeat the run exactly. (UNVERIFIED) simulation parameter set A simulation trajectory is the record of a molecule's motion over time, captured during a simulation. It consists of a time-ordered sequence of snapshots or "frames" where each frame contains the 3D coordinates of all atoms at a specific instant. This trajectory is the primary output of a molecular dynamics simulation and is the raw data used for all subsequent analysis of the system's dynamic behavior. trajectory simulation trajectory A simulation log is a text file that records everything that happened during a computer simulation. As a type of process output, the simulation log is a human-readable file that contains a time-ordered record of a simulation run, including an echo of input parameters, performance data, and periodic updates of thermodynamic properties. This file is the primary source for monitoring the progress and stability of a simulation and for diagnosing errors. simulation log A process output that holds the complete set of information needed to resume a simulation exactly where it stopped. It includes the atomic positions and velocities, the box dimensions, and the internal variables of any thermostat or barostat. Restarting from positions alone gives a different trajectory, because the missing quantities are re-initialised. (UNVERIFIED) simulation state Analysis data is the information content that is the result of analyzing simulation data. This includes any derived quantities calculated from a trajectory, such as RMSD values over time, a potential of mean force profile, or a list of identified hydrogen bonds. This processed data represents the scientific findings extracted from the simulation. analysis data A category for file formats designed to store information about molecules, such as their structure, movement, or the forces governing them. These formats provide standardized ways to represent molecular systems for use in computational chemistry and biology software. This class includes formats for atomic coordinates, trajectories, and force field parameters. molecular data format A prep format in which the geometry of the new building block is given as internal coordinates, that is, a bond length, an angle and a dihedral angle for each atom relative to atoms already placed. It is otherwise the same as the variant that uses Cartesian coordinates. Geometry in this form can define a building block's ideal shape without a starting 3-dimensional structure. (UNVERIFIED) AMBER prep (internal) format prep (internal) format OBSOLETE. A file format specifically designed to describe the complex, three-dimensional structures of large biological molecules like proteins and DNA. These formats accommodate hierarchical descriptions, such as chains and residues, in addition to atomic coordinates. Examples include the widely used PDB and mmCIF formats. Retired in the file-format flattening. Classifying structure formats by molecule size was an artificial distinction: a small molecule can be stored in a PDB file and a macromolecule in a mol2 file. Its former children are now direct children of 'molecular structure format' (MOLSIM:001088). Reported in GitHub issue #20. obsolete macromolecular structure format true A type of file format that describes the three-dimensional arrangement of atoms in a molecule. These formats store atomic coordinates and often include information about chemical bonds and residue identity. They are essential for visualizing molecules and providing the starting point for simulations. molecular structure format OBSOLETE. A coordinate file primarily stores the three-dimensional positions (X, Y, Z coordinates) of atoms in a molecular system. It may also contain atom names, residue information, and potentially periodic box dimensions, but typically lacks detailed bonding (topology) or force field parameter information. Common examples include PDB, GRO, and XYZ formats, used for initial structures or snapshots from simulations. Retired in the same file-format flattening. Grouping coordinate formats by the software that writes them proved unhelpful, since formats are not owned by one program. Its former children are now direct children of 'molecular structure format' (MOLSIM:001088). Related discussion in GitHub issue #20. obsolete software-specific coordinate format true A trajectory file stores the history of atomic positions (and sometimes velocities, forces, energies or the simulation box size) recorded at regular intervals during a molecular dynamics simulation. It contains a series of snapshots (or frames) representing the system's movement over time. Analyzing trajectories allows researchers to study dynamic processes, conformational changes, and calculate time-averaged properties. molecular trajectory format A file format that defines the atoms of a molecular system, how they are bonded together, and the force field parameters that apply to them. This includes atom types, partial charges, and parameters for bonds, angles, dihedral angles, and non-bonded interactions (like Lennard-Jones parameters). It essentially provides the 'rules' for how atoms interact in a molecular mechanics simulation. (UNVERIFIED) molecular topology format A prep format in which the geometry of the new building block is given as Cartesian coordinates, that is, 3 numbers per atom. It is otherwise the same as the variant that uses internal coordinates. Geometry in this form is easier to produce from an existing structure file. (UNVERIFIED) AMBER prep (cartesian) format prep (cartesian) format A text data format used by CHARMM that holds the numerical force field parameters: force constants, preferred bond lengths and angles, rotation terms and van der Waals values. It is paired with a separate file defining the residues those parameters apply to. The 2 together are what a CHARMM topology is built from. (UNVERIFIED) CHARMM parameter format par format CHARMM par format A data format that specifies the 3-dimensional arrangement of atoms inside a Gaussian input file, either as Cartesian x, y and z coordinates or as a Z-matrix of bond lengths, angles and dihedral angles. The Z-matrix form is chosen when internal coordinates or molecular symmetry need to be held fixed; the Cartesian form is simpler and more common. Both appear in the same place in the input file. (UNVERIFIED) GAUSSIAN coordinate format A data format specifying a molecule's geometry inside a Gaussian input file as bond lengths, angles and rotations relative to atoms already placed. It is chosen when a geometry must be scanned or constrained, because one number can be varied or frozen directly. It is harder to write than plain coordinates and is not used unless that control is wanted. (UNVERIFIED) GAUSSIAN z-matrix format A data format specifying a molecule's geometry inside a Gaussian input file as an element symbol followed by x, y and z coordinates for each atom. It is a simple and common way to give Gaussian a geometry, and is easy to generate from ordinary structure files. Unlike a Z-matrix, it does not single out bond lengths or angles that can be varied or held fixed. (UNVERIFIED) GAUSSIAN cartesian format A text data format holding the settings for a run of the paramfit program in AmberTools, which fits force field parameters to quantum mechanical or experimental reference data. It is written as assignments of the form variable=value, 1 per line, and says what kind of run is wanted, which parameters are to be fitted and which reference data to fit against. (UNVERIFIED) job control format A text data format holding the parameters of a ReaxFF reactive force field, which are far more numerous than a conventional force field needs because bonding is computed from the distances between atoms rather than fixed in advance. A single file covers every element combination the parameter set supports. Parameters fitted for different element sets are not interchangeable, so the file determines which chemistry can be simulated. (UNVERIFIED) ReaxFF force field format A data format, held in files known as Slater-Koster files, that stores the pre-calculated electronic and repulsive properties between pairs of atom types for a density functional tight binding calculation. The values are derived once from more expensive quantum mechanical calculations and then reused. That reuse is what makes the method fast while keeping a quantum mechanical description of the system. (UNVERIFIED) skf format DFTB parameter format A text data format listing the covalent bonds that may be cut when the core, quantum and buffer regions of an adaptive buffered force-mixing QM/MM simulation are drawn, so that region boundaries fall at chemically sensible places. Each line names 2 atoms, by atom type or by index, joined by an arrow that says from which side the bond may be broken. (UNVERIFIED) cuttable bond list format A text data format that defines the clusters used in reservoir replica exchange molecular dynamics with dihedral bin clustering, listing the dihedral angles that define the clusters, the weight of each cluster and which cluster each reservoir structure belongs to. The weights let structures from the reservoir enter the simulation with a non-Boltzmann weighting. It can be written by the dihedral clustering analysis of cpptraj. (UNVERIFIED) reservoir cluster info file format A simulation parameter format that holds the control settings for an EMIL absolute free energy calculation in AMBER, conventionally in a file named emilParameters.in. It records the reference model parameters and the integration settings that the calculation needs. (UNVERIFIED) Berryman J.T. Absolute free energies for biomolecules in implicit or explicit solvent. Physics Procedia, 2014. States that the AMBER implementation of this method is called EMIL. Berryman J.T. and Schilling T. Free energies by thermodynamic integration relative to an exact solution, used to find the handedness-switching salt concentration for DNA. J. Chem. Theory Comput. 9(1):679-686, 2013. emilParameters.in format EMIL input format The cein format is a specialized input file used in constant redox-potential MD simulations. It defines the properties of residues that can exist in different oxidation states, such as a cysteine that can be oxidized or reduced. This file provides the necessary parameters for a constant redox potential simulation, allowing the system to dynamically change its electronic state in response to a defined electrochemical potential. cein format The cpein format is a complex input file for constant pH and electrochemical potential (CpH/E) MD simulations that simultaneously model changes in both protonation and electron-transfer states. It defines the properties of residues that can be affected by both the surrounding pH and the electrochemical potential. This format is required when studying complex bioenergetic processes where proton and electron transfers are coupled. cpein format A molecular structure format, based on the macromolecular Crystallographic Information File format, that describes 1 chemical building block, giving its atom names, elements, bonds, charges and ideal coordinates. Each file acts as a dictionary entry that deposited structures refer to rather than repeat, as in the Chemical Component Dictionary of the Protein Data Bank. It is how a ligand or a modified residue is described once and reused. (UNVERIFIED) component cif A data format that stores the results of analysing how solvent molecules are arranged and behave around a solute. Files in these formats typically hold values on a 3-dimensional grid, such as the chance of finding a solvent atom at a point or the average interaction energy there. Examples include solvent density maps and the correlation functions produced by 3D-RISM theory. (UNVERIFIED) solvent analysis data format A data format that stores a 3-dimensional grid giving the chance of finding a solvent molecule at each position around a solute, relative to a random distribution. The value at each point is the pair distribution function g(r): above 1 means the solvent gathers there, below 1 means it is pushed away. Peaks in the grid mark solvation shells and hydrogen bonding sites. (UNVERIFIED) solvent pair distribution format A data format that stores a 3-dimensional grid of the direct correlation function c(r) between a solute and the surrounding solvent. It isolates the solute's direct influence on a solvent molecule at each position, leaving out effects passed along through other solvent molecules. It is a basic quantity in liquid-state theories such as 3D-RISM and is used to derive thermodynamic properties of solvation. (UNVERIFIED) solvent direct correlation format A data format that stores a 3-dimensional grid of the total correlation function h(r) between a solute and the surrounding solvent. It gives the whole departure from a random solvent distribution caused by the solute, counting both direct and indirect effects, and it relates to the pair distribution function as h(r) = g(r) − 1. It is used in liquid-state theories to calculate solvation properties. (UNVERIFIED) solvent total correlation format This file format stores a 3D map of the potential energy experienced by the solvent molecules in the space surrounding a central solute. The value at each grid point represents the average potential energy of the solvent at that location, reflecting the combined interactions with the solute and other solvent molecules. This format provides a view of the energetic landscape of the solvent, which can be used to identify energetically favorable or unfavorable regions for solvation. solvent potential file format A data format that stores a 3-dimensional grid of the average electric charge density contributed by solvent molecules around a solute. Unlike a map of where solvent atoms sit, it gives the net positive or negative charge at each point, averaged over the simulation. It shows how the solvent screens and responds to the solute's own charges. (UNVERIFIED) solvent charge density distribution format A data format that stores a 3-dimensional grid giving the solvent charge at each grid point around a solute, as a point charge in units of the elementary charge, combining the contributions of all solvent atom types. It differs from a charge density map, which gives charge per unit volume at each point. It is an output of 3D-RISM calculations. (UNVERIFIED) solvent charge distribution format A data format that stores a 3-dimensional grid of the interaction energy a solvent molecule would feel at each position around a solute. Negative values mark positions the solvent is drawn to and positive values mark positions it is pushed away from. The grid shows the energetic landscape that determines how solvent arranges itself. (UNVERIFIED) solute-solvent potential energy format A text data format in which a 1D-RISM calculation writes how a property of the bulk solvent, such as the pair distribution, the direct correlation or the site-site potential, varies with distance for every pair of solvent sites. After a few header lines starting with #, each line gives a separation followed by the value for each site pair. The susceptibility file that 3D-RISM reads is written in a different format. (UNVERIFIED) site-site functionals format A data format that stores, for every pair of sites in a bulk solvent, the total excess coordination number: how many more sites of one type surround a site of the other type than a uniform distribution would give. It measures how strongly the solvent structures itself. It is written by 1D-RISM calculations in a file ending in .n00. (UNVERIFIED) total excess coordination number format A data format that stores, for every pair of sites in a bulk solvent, the total excess charge of one site type gathered around the other, compared with a uniform distribution. It is laid out as a table with 1 row and 1 column per solvent site, plus the total excess charge from all sites. It is written by 1D-RISM calculations in a file ending in .q00. (UNVERIFIED) total excess charge format A data format in which Gaussian writes the electrostatic potential it computed at a set of points around a molecule, together with those points' coordinates. It is the input a charge fitting program reads. The choice of where the points lie affects the charges obtained, so it is part of the fitting protocol rather than an incidental detail. (UNVERIFIED) GAUSSIAN electrostatic potential format The mdinfo format is a text file generated by the Amber simulation software that provides a concise, real-time summary of a running simulation. It contains the most recently calculated energy terms and thermodynamic properties, such as temperature and pressure, formatted for easy monitoring. This file is continuously updated during a run, making it useful for checking the immediate status and stability of a simulation without having to parse the much larger main output file. AMBER mdinfo format mdinfo format A database web server is an online application that provides a graphical interface for searching, visualizing, and retrieving data from a structured repository. It acts as an accessible gateway, allowing researchers to query complex datasets and download specific entries without needing direct interaction with the underlying database management system. database web server A standard text-based file format designed for exchanging NMR-derived information, particularly structural restraints like distances and angles. It provides a standardized way to represent experimental data used for structure calculation and simulation refinement. (UNVERIFIED) NMR exchange format NEF format An input variable or algorithmic setting defined prior to execution that governs the behavior, conditions, or numerical rules of a computational process (encompassing both simulation generation and post-run analysis). (UNVERIFIED) parameter An input setting that determines the physical interactions, boundary conditions or integration rules used to generate a trajectory. Examples are the target temperature a thermostat holds, the size of the timestep, and the distance beyond which interactions are ignored. These settings are part of what a result means, so a simulation cannot be reproduced or judged without them. (UNVERIFIED) simulation parameter A simulation parameter controlling how the equations of motion are advanced in time, including the timestep, the integration algorithm and how long the run lasts. The timestep matters most: too large a value makes the simulation unstable, and too small a value makes it slower than necessary. (UNVERIFIED) integration parameter An integration parameter that gives the length of simulated time between 2 successive steps of a molecular dynamics integration. It must be short enough to follow the fastest atomic motions, such as the vibration of bonds to hydrogen, for the integration to stay stable, while a longer step lowers the cost of covering a given length of simulated time. (UNVERIFIED) simulation time step An integration parameter that gives the total length of simulated time for which a simulation is run. It determines how many different conformations can be visited and which slow processes can be observed at all. In most programs it is set indirectly, as a number of steps multiplied by the time step. (UNVERIFIED) simulation duration A specific type of simulation duration that defines the length of a single stage within a multi-stage simulation protocol. This is used in methods like steered MD or umbrella sampling but also during density adaptation and equilibration workflows, where the overall process is broken down into smaller, sequential runs. The simulation time per stage ensures that each part of the protocol is run for a defined period. This is often specified as nsteps within a block for a specific stage. simulation time per stage An integration parameter giving the conditions under which an energy minimization stops. The usual conditions are that the largest force on any atom, or the root-mean-square of all the forces, has fallen below a threshold, or that the energy has stopped changing between steps. A minimization that stops too early leaves strain in the structure, which can make the first steps of the following simulation unstable. (UNVERIFIED) minimization convergence criteria A minimization convergence criterion based on the root-mean-square of the energy gradient, which measures how large the forces on the atoms are. The minimization stops when this root-mean-square value falls below the chosen threshold. It is a common way to define when a minimization has converged. (UNVERIFIED) AMBER 26 Reference Manual, p. 434, entry drms. minimization drms criterion A minimization convergence criterion giving the threshold on the energy gradient below which an energy minimization is stopped. A smaller value makes the geometry optimization stricter and needs more steps. (UNVERIFIED) minimization gradient tolerance A simulation parameter that fixes which quantities are held constant during a run, and so which statistical ensemble is sampled. Holding particle number, volume and temperature gives one ensemble; holding pressure instead of volume gives another. The choice determines which averages taken from the trajectory are meaningful. (UNVERIFIED) thermodynamic ensemble parameter A thermodynamic ensemble parameter specifying how the temperature is held at its target, covering the thermostat algorithm, the target value and the coupling strength. A thermostat both holds the temperature and changes the dynamics, so these settings affect more than the average temperature. They must be reported with any result that depends on rates. (UNVERIFIED) temperature control parameter A temperature control parameter naming which algorithm holds the temperature, such as Berendsen, Nosé-Hoover or stochastic velocity rescaling. They differ in whether they produce the correct spread of energies or only the correct average. The Berendsen scheme does not, which is why it is used to bring a system to temperature and then replaced. (UNVERIFIED) thermostat algorithm parameter A temperature control parameter that specifies the temperature at which the simulation is to be held. The thermostat adjusts the velocities of the atoms so that the instantaneous temperature fluctuates around this value. It is used only when a thermostat is applied, not in a run at constant energy. (UNVERIFIED) AMBER 26 Reference Manual, p. 436, entry temp0. target temperature A temperature control parameter setting how strongly the system is coupled to the simulated heat bath. Strong coupling holds the temperature close to target and interferes with the natural motion; weak coupling leaves the dynamics alone and lets the temperature wander. Any quantity involving rates depends on this choice. (UNVERIFIED) thermostat coupling parameter A thermostat coupling parameter specifying the time constant of the coupling between the system and the heat bath, which sets how quickly the kinetic temperature is driven towards the target temperature. AMBER 26 Reference Manual, p. 436, entry tautp. thermostat coupling constant temperature coupling time constant A thermostat coupling parameter giving how often, on average, a particle receives a random kick from the simulated bath in a stochastic thermostat. A high frequency reaches the target temperature quickly and turns the motion into a random walk; a low one preserves the real dynamics and controls temperature weakly. It is the stochastic counterpart of a coupling time constant. (UNVERIFIED) AMBER 26 Reference Manual, p. 216, entry gamma_ln. thermostat collision frequency A thermodynamic ensemble parameter specifying how the pressure is held at its target, covering the barostat algorithm, the target value, the coupling strength and the compressibility. The barostat changes the volume of the simulation box so that the average pressure matches the target. These settings apply only to constant-pressure runs. (UNVERIFIED) pressure control parameter A pressure control parameter naming which algorithm holds the pressure at its target, such as Berendsen, Parrinello-Rahman or a Monte Carlo scheme. They differ in whether they generate the correct distribution of volumes or merely the correct average. The Berendsen scheme equilibrates quickly and does not give the correct distribution, which is why it is used to settle a system and then replaced. (UNVERIFIED) barostat algorithm parameter A pressure control parameter that specifies the pressure at which a constant-pressure simulation is to be held. The barostat changes the volume of the simulation box so that the instantaneous pressure fluctuates around this value. It is used only in constant-pressure simulations. (UNVERIFIED) AMBER 26 Reference Manual, p. 438, entry pres0. target pressure A pressure control parameter giving the time constant over which a barostat relaxes the system pressure towards the target pressure. AMBER 26 Reference Manual, p. 438, entry taup. barostat coupling constant pressure coupling time constant A pressure control parameter giving how much the system's volume changes for a given change in pressure. Some barostats need it to work out how far to scale the box when correcting pressure. For a system in water the value for water is normally used, because water fills most of the box. (UNVERIFIED) AMBER 26 Reference Manual, p. 438, entry comp. compressibility A simulation parameter controlling how interactions between atoms are computed, covering the cutoff distances, the long-range electrostatic method and the treatment at the cutoff. These settings determine how much of each interaction is actually included. They affect both cost and correctness, which is why they are reported with any published simulation. (UNVERIFIED) interaction parameter An interaction parameter controlling how van der Waals and short-range electrostatic interactions are computed, covering the cutoff distance, the treatment at that cutoff, and the neighbour list settings. These settings decide how much of the interaction is included and what happens at the boundary. They are among the few settings that affect both speed and correctness directly. (UNVERIFIED) non-bonded interaction parameter A non-bonded interaction parameter naming how interactions are handled at the cutoff distance, such as truncating them sharply, shifting them so that they reach zero at the cutoff, or smoothly switching them to zero over a range. A sharp truncation makes the energy and the force change abruptly when a pair of atoms crosses the cutoff, which harms energy conservation. The choice therefore affects the stability of a simulation as well as its accuracy. (UNVERIFIED) non-bonded method parameter A non-bonded interaction parameter that specifies the single distance beyond which both van der Waals and electrostatic interactions are no longer calculated directly. Leaving out pairs farther apart than this distance is one of the main ways to reduce the cost of a simulation, and using 1 distance for both kinds of interaction is common. Programs that allow the 2 distances to be set separately use a more specific parameter for each. (UNVERIFIED) AMBER 26 Reference Manual, p. 456, entry cut. AMBER 26 Reference Manual, p. 467, card CUT, which names this quantity the non-bonded cutoff distance. NAMD 3.x User's Guide, p. 55, section 5.2.3, entry cutoff. States that the cutoff is the local interaction distance common to both electrostatic and van der Waals calculations. cut non-bonded cutoff distance cut A non-bonded interaction parameter giving the distance at which a switching function starts to reduce the interaction, so that it reaches zero at the cutoff. Without such smoothing, the energy and the force change abruptly when a pair of atoms crosses the cutoff, which harms energy conservation. The gap between this distance and the cutoff sets how gradual that transition is. (UNVERIFIED) switching distance A non-bonded interaction parameter giving how many steps pass between rebuilds of the neighbour list. Rebuilding often is safe and costly; rebuilding rarely is cheap and risks missing a pair that has moved into range. It is chosen together with the pairlist skin, since a thicker skin allows a longer interval. (UNVERIFIED) AMBER 26 Reference Manual, p. 456, entry nsnb. pairlist update frequency A non-bonded interaction parameter giving the extra distance beyond the cutoff within which pairs of atoms are recorded in the neighbour list. Including pairs that are not yet interacting means the list stays valid for several steps as atoms move, so it need not be rebuilt every step. Too small a value lets a pair reach interaction range unnoticed, which silently loses interactions. (UNVERIFIED) pairlist skin An interaction parameter controlling how electrostatic interactions are computed. Its members cover which long-range method is used and the settings that method needs, such as grid spacing and real-space cutoff. These settings determine accuracy and cost together, so they are chosen as a set rather than individually. (UNVERIFIED) electrostatic parameter An electrostatic parameter naming which algorithm computes the long-range electrostatic interactions, such as particle mesh Ewald, plain Ewald summation, or a reaction field. Electrostatic forces fall off slowly enough that simply ignoring them past a cutoff introduces visible artefacts. The choice of method is therefore part of what a published simulation must state. (UNVERIFIED) long-range electrostatic method parameter An electrostatic parameter giving how far apart the points of the reciprocal-space grid lie in a particle mesh Ewald calculation. A finer grid is more accurate and costs more, and the cost rises with the cube of the refinement. It trades off against the real-space cutoff, so the 2 should be chosen together rather than separately. (UNVERIFIED) AMBER 26 Reference Manual, p. 500, entry fft_grids_per_ang. PME grid spacing An electrostatic parameter used by the particle mesh Ewald (PME) method that gives the number of grid points along each box dimension for the reciprocal-space calculation. More grid points for the same box give a more accurate calculation at a higher cost. The numbers are usually chosen automatically from the box size and a target grid spacing, but can also be set directly. (UNVERIFIED) PME grid dimensions An electrostatic parameter giving how strongly a medium screens the electrostatic interaction between charges within it, relative to a vacuum. It is 1 in vacuum and usually also in explicit solvent simulations, where the water molecules do the screening themselves. Implicit solvent models instead give the inside of the solute a low value and the surrounding solvent a high one, to stand in for solvent that is not present. (UNVERIFIED) dielectric constant A dielectric constant that describes the interior of the solute in a continuum solvation calculation. It is set low, usually between 1 and 4, because the packed interior of a molecule screens charge poorly. Its value is one of the largest sources of variation between published binding free energy estimates. (UNVERIFIED) AMBER 26 Reference Manual, p. 82, entry epsin. AMBER 26 Reference Manual, p. 952, entry indi. internal dielectric constant A dielectric constant that describes the bulk solvent surrounding a solute in a continuum solvation calculation. It is set high, near 80 for water, so that the solvent screens electrostatic interactions strongly. Together with the internal value it fixes the dielectric boundary at the molecular surface. (UNVERIFIED) AMBER 26 Reference Manual, p. 82, entry epsout. AMBER 26 Reference Manual, p. 952, entry exdi. external dielectric constant An electrostatic parameter giving the distance at which the directly summed part of an Ewald calculation stops. Raising it moves work out of the reciprocal-space part and into the real-space part, and lowering it does the reverse. The accuracy of the whole calculation depends on this distance together with the reciprocal-space grid and the coefficient that sets how the interaction is divided between the 2 parts. (UNVERIFIED) AMBER 26 Reference Manual, p. 499, entry es_cutoff. direct sum cutoff An interaction parameter giving the strength, direction and time dependence of a field applied to the system from outside. It is what specifies an applied electric field in a simulation of ion transport or of a membrane under voltage. Without it the field has no defined magnitude. (UNVERIFIED) AMBER 26 Reference Manual, p. 457, entry efx. external field parameter An external field parameter that gives the strength of an applied external electric field along 1 Cartesian axis. The components along the 3 axes together set the direction and strength of a uniform field acting on the charged particles of the system. Such a field is used to simulate the effect of an applied field or voltage on the molecules. (UNVERIFIED) external electric field component A type of simulation parameter that controls the application of geometric constraints or restraints to the system. Constraints are used to fix certain degrees of freedom, while restraints are used to guide the system towards a particular conformation. These constraint and restraint parameter settings are used to control the flexibility of the system or to apply experimental data. constraint and restraint parameter A parameter controlling how geometric constraints are applied, most often to hold bonds involving hydrogen at a fixed length. Fixing those bonds removes the fastest motion in the system, which permits a longer timestep, typically about twice as long. Its members say which bonds are held, by which algorithm, and how exactly. (UNVERIFIED) constraint parameter A constraint parameter naming which algorithm enforces the geometric constraints, such as SHAKE, RATTLE, LINCS or SETTLE. The choice affects speed and how well the constraints are held, not what is being constrained. SETTLE is used for water specifically, because water's constraints have an exact solution that needs no iteration. (UNVERIFIED) constraint algorithm parameter A constraint parameter naming which bonds are held at a fixed length, usually as a choice between all bonds, only the bonds involving hydrogen, or none. Constraining only the bonds involving hydrogen is the most common choice, because most force fields were parameterised that way and it permits a longer timestep. Constraining all bonds is slower and is not needed at timesteps up to about 2.5 femtoseconds. (UNVERIFIED) constrained bonds constrained bond A constraint parameter giving how close to its target length a constrained bond must be before the algorithm stops adjusting it. A tighter value holds the geometry more exactly and takes more iterations each step. Too loose a value lets energy drift over a long run, which is the usual sign it was set badly. (UNVERIFIED) SHAKE tolerance A type of constraint and restraint parameter that specifies how artificial restraining potentials are applied to the system. These are used to guide the simulation towards a specific conformation. The restraint parameter settings are essential for biased simulations and structure refinement. These are often set with keywords like restraint_wt and restraintmask. restraint parameter A restraint parameter that specifies the force constant for a harmonic or other restraining potential. A larger restraint energy constant results in a stronger force pulling the restrained atoms towards their target values. This is a key parameter for controlling the strength of the applied restraints. This is often set with the keyword restraint_wt. AMBER 26 Reference Manual, p. 433, entry restraint_wt. restraint energy constant A restraint parameter that names which atoms a restraining force applies to, written as a selection expression rather than a list. It is what lets part of a system be held in place while the rest moves freely, such as holding a protein still while water settles around it. A mask that selects the wrong atoms produces a simulation that runs normally and means something else. (UNVERIFIED) AMBER 26 Reference Manual, p. 433, entry restraintmask. restraint mask A specific restraint energy constant used to apply a restraining potential to a spherical cap of water molecules around a solute. This is used in some simulation setups to prevent the solute from becoming exposed to the vacuum. The water cap force constant controls the strength of this spherical boundary potential. This is often set with the keyword fcap. AMBER 26 Reference Manual, p. 454, entry fcap. water cap force constant A simulation parameter that fixes what the system is made of and what state it is in, rather than how it is advanced in time. Examples are the box dimensions, the pH and the protonation states, the ionic strength and the state of the chain ends. Most are chosen once while the system is built, although some, such as the box size in a constant-pressure run, change while it runs. (UNVERIFIED) system setup parameter A system setup parameter giving the acidity of the simulated solution. It determines which acidic and basic groups carry a proton, and so the charge on the molecule. In an ordinary simulation it is used once to assign fixed protonation states; in a constant pH simulation it is held during the run and the states respond to it. (UNVERIFIED) pH A system setup parameter that specifies the molar concentration of salt in the simulation solvent. The ionic strength affects the screening of electrostatic interactions and is crucial for mimicking experimental buffer conditions. This is a key parameter for accurately modeling the electrostatic environment of the system. This is often set with the keyword ion_concentration or saltcon. AMBER 26 Reference Manual, p. 73, entry saltcon. AMBER 26 Reference Manual, p. 984, entry ion_concentration. ionic strength A system setup parameter that specifies the size and shape of the periodic box used to create a periodic system. The periodic box dimensions are essential for any simulation using periodic boundary conditions. They determine the overall density of the system. periodic box dimensions A system setup parameter giving the chemical state of each end of a polymer chain: charged, neutral, or capped with a group that mimics the continuation of the chain. Capping is used when the simulated fragment is part of a longer chain in reality, so that an artificial charge is not introduced. A wrong choice adds or removes charge at exactly the place a fragment is most exposed. (UNVERIFIED) molecular termini state A system setup parameter that specifies whether the simulation involves an alchemical transformation for a free energy calculation. This parameter indicates that some atoms/residues in the system have their properties changed over the course of the simulation. The alchemical transformation state is a key descriptor for identifying free energy calculations. This is often controlled by a perturbation flag (lambda, λ). alchemical transformation state A simulation process parameter that specifies the number of independent simulation runs performed from the same or similar initial conditions. Running number of replicates helps to assess the statistical robustness and convergence of simulation results. This is a key parameter of the overall simulation experimental design. number of replicates A system setup parameter giving the target value of the tension, a force per unit length, acting in the plane of an interface in a constant-pressure simulation of a membrane or a liquid surface. The box is scaled in the plane of the interface to bring the tension towards this value, while the pressure normal to the interface is controlled separately. A value of 0, the AMBER default, describes an interface under no tension. (UNVERIFIED) AMBER 26 Reference Manual, p. 439, entry gamma_ten. surface tension A system setup parameter giving the number of QM/MM partitionings used in an adaptive solvent QM/MM simulation, in which solvent molecules can pass between the quantum and the classical region. Each partitioning assigns the solvent molecules in a transition region between the 2 regions differently to the quantum or the classical description, and the energy is a weighted average over the partitionings. With N solvent molecules in the transition region, N + 1 partitionings carry a nonzero weight. (UNVERIFIED) AMBER 26 Reference Manual, p. 211, entry n_partition. number of QM/MM partitions A type of simulation parameter that controls the application of a specific enhanced sampling method designed to accelerate the exploration of the system's conformational space. These parameters define the nature and strength of the biasing potential or other modifications to the standard simulation protocol. enhanced sampling parameter A type of enhanced sampling parameter that controls a steered molecular dynamics (sMD) simulation, where an external force is applied to pull the system along a specific coordinate. These parameters define the pulling speed, force constant, and direction of the steering. The steered molecular dynamics parameter settings are essential for setting up non-equilibrium pulling simulations. These are often set with keywords like smd. steered molecular dynamics parameter A steered molecular dynamics parameter that specifies the constant speed at which the steered coordinate is changed during a constant-velocity sMD simulation. The atom pulling velocity determines the rate at which the system is pulled in agreement with the selected reaction coordinate. This is often set with the keyword v. atom pulling velocity A steered molecular dynamics parameter that specifies the stiffness of the harmonic potential used to pull the system along the reaction coordinate. A larger steered MD restraint energy constant results in a stiffer spring and less deviation from the target path. This is a key parameter for controlling the strength of the steering. This is often set with the keyword harm. AMBER 26 Reference Manual, p. 610, entry harm. steered MD restraint energy constant A steered molecular dynamics parameter that specifies the total number of distinct stages or windows used in a staged sMD simulation protocol. This allows for the pulling to occur incrementally or for parameters to be changed between stages. The number of sMD stages is a key parameter for setting up complex pulling protocols. This is often set with the keyword nstage. number of sMD stages A type of enhanced sampling parameter that controls an umbrella sampling simulation, where a biasing potential is used to confine the system to specific windows along a reaction coordinate. These parameters define the location and strength of the biasing potential in each window. The umbrella sampling parameter settings are essential for setting up simulations to calculate a potential of mean force. These are often set in the input file for each window. umbrella sampling parameter An umbrella sampling parameter that specifies the strength of the harmonic biasing potential applied in each umbrella sampling window. This force constant determines how strongly the system is restrained around the target value of the reaction coordinate for that window. The umbrella sampling restraint energy constant is a key parameter for ensuring sufficient sampling within each window. This is often set with the keyword anchor_strength. umbrella sampling restraint energy constant An umbrella sampling parameter that specifies the total number of individual simulations (windows) performed in an umbrella sampling workflow. Each window is typically centered at a different value along the reaction coordinate. The number of umbrella sampling windows is a key parameter for ensuring adequate coverage of the entire reaction coordinate. number of umbrella sampling windows An umbrella sampling parameter that determines the center or boundaries of the harmonic potential for a specific umbrella sampling window. This parameter defines the region of the reaction coordinate that will be sampled in that window. The umbrella anchor position is a key parameter that is varied between different windows. This is often set with the keyword anchor_position. umbrella anchor position A type of enhanced sampling parameter that controls an accelerated molecular dynamics (aMD) simulation. These parameters define the energy threshold and strength of the boost potential that is applied to flatten the energy landscape. The accelerated MD parameter settings are crucial for controlling the degree of enhanced sampling. These are often set in a specific amd section of the input file. accelerated MD parameter An accelerated MD parameter that specifies the energy threshold for the dihedral energy, below which the biased potential is applied. This parameter controls when the dihedral energy term is biased. The aMD dihedral energy threshold is a key parameter for tuning the aMD simulation. This is often set with the keyword EthreshD. aMD dihedral energy threshold A scaling parameter used to flatten high potential energy barriers, allowing the simulation to explore conformational space much faster. It dictates the magnitude of the bias applied to the system's dihedral degrees of freedom. A smaller aMD dihedral alpha factor results in a larger boost. This is a key parameter for tuning the aMD simulation. This is often set with the keyword alphaD. aMD dihedral alpha factor An accelerated MD parameter that specifies the energy threshold for the total potential energy, below which the biased potential is applied. This parameter controls when the total potential energy is biased. The aMD potential energy threshold is a key parameter for tuning the aMD simulation. This is often set with the keyword EthreshP. aMD potential energy threshold An accelerated MD parameter that determines the strength of the biased potential applied to the total potential energy term. A smaller aMD potential alpha factor results in a larger boost. This is a key parameter for tuning the aMD simulation. This is often set with the keyword alphaP. aMD potential alpha factor A type of enhanced sampling parameter that controls a Gaussian accelerated molecular dynamics (GaMD) simulation. These parameters define the upper limit of the standard deviation for the potential energy boost. The Gaussian accelerated MD parameter settings are crucial for controlling the balance between accuracy and enhanced sampling. These are often set in a specific gamd section of the input file. Gaussian accelerated MD parameter A Gaussian accelerated MD parameter that specifies the upper limit of the standard deviation for the potential energy boost. This parameter controls the maximum strength of the applied bias. This is often set with keywords like sigma0P or sigma0D. AMBER 26 Reference Manual, p. 526, entry sigma0D. AMBER 26 Reference Manual, p. 526, entry sigma0P. GaMD potential boost stddev limit A type of enhanced sampling parameter that controls a random acceleration molecular dynamics (RAMD) simulation. These parameters define the strength of the randomly applied acceleration force. The random acceleration MD parameter settings are crucial for enhancing the exploration of conformational space. These are often set in a specific ramd section of the input file. random acceleration MD parameter A random acceleration MD parameter that specifies the strength of the acceleration boost applied in a Random Acceleration Molecular Dynamics (RAMD) simulation. This parameter controls the magnitude of the random forces used to accelerate the system's exploration. The RAMD boost strength is a key parameter for tuning the RAMD simulation. This is often set with the keyword ramdboost. AMBER 26 Reference Manual, p. 458, entry ramdboost. RAMD boost strength A type of enhanced sampling parameter that controls a targeted molecular dynamics (TMD) simulation. These parameters define the target structure and the strength of the biasing potential used to guide the simulation towards it. The targeted MD parameter settings are essential for simulating large-scale conformational transitions between known start and end states. These are often set in a specific tmd section of the input file. targeted MD parameter A targeted MD parameter that specifies the force constant for the RMSD restraint potential used to guide the simulation towards the target structure. A larger targeted MD force constant results in a stronger biasing force. This is a key parameter for controlling the strength of the targeted MD simulation. This is often set with the keyword tgtmdfrc. AMBER 26 Reference Manual, p. 532, entry tgtmdfrc. targeted MD force constant A targeted MD parameter that specifies the target root-mean-square deviation (RMSD) value for the restraint. The simulation is guided to reduce the RMSD of the current structure relative to a reference structure to this targeted MD RMSD target value. This is a key parameter for defining the endpoint of the TMD simulation. This is often set with the keyword tgtrmsd. AMBER 26 Reference Manual, p. 532, entry tgtrmsd. targeted MD RMSD target A type of enhanced sampling parameter that controls a nudged elastic band (NEB) simulation, which is used to find the minimum energy path between two conformations. These parameters define the number of intermediate structures (images) and the spring constants connecting them. The nudged elastic band parameter settings are essential for accurately mapping reaction pathways. These are often set in a specific neb section of the input file. nudged elastic band parameter A nudged elastic band parameter that specifies the spring force constant connecting adjacent images along the reaction path. This parameter ensures that the images are evenly spaced and do not drift off the path. The NEB spring constant is a key parameter for the stability and convergence of the NEB calculation. This is often set with the keyword nebk. AMBER 26 Reference Manual, p. 550, entry nebk. NEB spring constant A nudged elastic band parameter that specifies the number of intermediate structures (images) to optimize along the reaction path. A larger NEB number of images provides a more detailed representation of the path but is computationally more expensive. This is a key parameter for controlling the resolution of the NEB calculation. This is often set with the keyword neb_nimage. AMBER 26 Reference Manual, p. 550, entry neb_nimage. NEB number of images A type of enhanced sampling parameter that controls a well-tempered metadynamics simulation. These parameters define how the height of the added Gaussian hills is rescaled over time, ensuring that the simulation eventually converges. The well-tempered metadynamics parameter settings are crucial for obtaining converged free energy surfaces. These are often set in a specific metadynamics section of the input file. well-tempered metadynamics parameter A well-tempered metadynamics parameter that controls the rate at which the height of the added Gaussian hills decreases. A higher well-tempered pseudo-temperature results in a faster convergence but a less detailed exploration of the energy landscape. This is a key parameter for tuning the well-tempered metadynamics simulation. This is often set with the keyword wt_temperature. AMBER 26 Reference Manual, p. 613, entry wt_temperature. well-tempered pseudo-temperature A simulation parameter that controls a calculation in which part of the system is treated with quantum mechanics and the rest with a force field. Its members set which atoms are in the quantum region, which quantum method is used, and how bonds crossing the boundary are handled. Such hybrid calculations are used mainly to study chemical reactions, such as those in enzyme active sites, where bonds are formed or broken. (UNVERIFIED) QM/MM parameter A QM/MM parameter giving the total charge of the quantum region as a whole number. The quantum calculation needs a definite number of electrons, so this cannot be left to be worked out. Setting it wrong gives a calculation that runs to completion and describes a different chemical species. (UNVERIFIED) AMBER 26 Reference Manual, p. 159, entry qmcharge. QM charge A QM/MM parameter giving the distance from a quantum atom to the link atom placed along its bond to a classical atom. The link atom, usually a hydrogen, fills the free valence left where the boundary between the quantum and classical regions cuts a covalent bond. In AMBER all link atoms are placed at the same distance, by default 1.09 ångström. (UNVERIFIED) AMBER 26 Reference Manual, p. 175, entry lnk_dis. QM/MM link atom distance A QM/MM parameter giving the cutoff distance for electrostatic interactions between the quantum region and the classical region. Classical atoms within this distance of any quantum atom interact directly with the quantum region, and when a particle mesh Ewald treatment is used the value only divides the work between the direct and the reciprocal parts. In AMBER it is by default equal to the cutoff used for the classical region. (UNVERIFIED) AMBER 26 Reference Manual, p. 172, entry qmcut. QM/MM non-bonded cutoff A QM/MM parameter controlling a density functional tight binding calculation, covering which parameter files are used, how many charge iterations are allowed and how tightly they must converge. The parameter files are specific to each pair of elements, so a system containing an unparameterised pair cannot be run. Checking that first saves discovering it at run time. (UNVERIFIED) DFTB parameter A parameter giving the greatest number of self-consistent charge iterations a tight binding calculation may take. It stops a calculation that will not converge from running indefinitely. Reaching the limit means the charges are not converged and the result must not be used. (UNVERIFIED) AMBER 26 Reference Manual, p. 159, entry dftb_maxiter. DFTB max iterations A DFTB parameter giving the electronic temperature used to spread the occupation of orbitals near the highest occupied level according to a Fermi distribution, which helps the self-consistent charge iterations of a density functional tight binding calculation converge. The spreading mixes the highest occupied and lowest unoccupied orbitals a little, which helps in difficult cases where the occupation would otherwise switch between iterations. It is a numerical aid rather than a physical temperature, so it is used only when needed and the results are then checked carefully. (UNVERIFIED) AMBER 26 Reference Manual, p. 159, entry dftb_telec. DFTB electronic temperature A QM/MM parameter controlling the self-consistent field procedure, the iterative calculation that finds the electronic structure of the quantum region. Its members set how closely successive iterations must agree, how many iterations are allowed, and which scheme is used to speed up or stabilise convergence. They are usually left at their defaults and adjusted when a difficult system fails to converge. (UNVERIFIED) SCF parameter A self-consistent field parameter giving the greatest number of iterations the calculation may take before it is treated as not converged. It stops a calculation that will not converge from running indefinitely. A result reached at the limit is not converged, although some programs continue with it, so it needs checking before it is used. (UNVERIFIED) AMBER 26 Reference Manual, p. 161, entry itrmax. max SCF iterations A parameter giving how little the energy or the electron density must change between iterations for a self-consistent field calculation to be considered finished. A tighter value gives a more accurate result and takes more iterations. Forces need a tighter setting than energies, because they are more sensitive to a partly converged density. (UNVERIFIED) AMBER 26 Reference Manual, p. 160, entry scfconv. SCF convergence criteria A parameter giving how much of the previous iteration's electron density is mixed into the next one. Mixing slows the calculation down and stops it oscillating between 2 near-solutions. It is raised when a difficult system will not converge and lowered otherwise, because it costs iterations. (UNVERIFIED) AMBER 26 Reference Manual, p. 479, entry damp. SCF damping factor A self-consistent field parameter that controls the direct inversion in the iterative subspace procedure, which speeds up convergence by extrapolating from previous iterations. The values set how many previous iterations are kept and how many extrapolation attempts are made. Poor settings make a calculation converge slowly or oscillate without settling. (UNVERIFIED) DIIS parameter A DIIS parameter giving the number of self-consistent field iterations in which an extrapolation by direct inversion in the iterative subspace is attempted. In AMBER these attempts start only after 100 iterations have failed to converge, so the setting matters only for molecules that are hard to converge. A larger value gives the extrapolation more iterations in which to succeed. (UNVERIFIED) AMBER 26 Reference Manual, p. 162, entry ndiis_attempts. DIIS attempts A DIIS parameter giving how many previous error vectors the convergence accelerator keeps and extrapolates from. It sets the size of the subspace used, so a larger value uses more history and more memory. Too few gives a poor extrapolation and too many can include iterations no longer relevant. (UNVERIFIED) ndiis_matrices DIIS matrices number of DIIS matrices ndiis_matrices A simulation parameter controlling a constant pH run, covering how often protonation changes are attempted, how long the system is allowed to relax after one, and at what pH the simulation is held. These settings determine whether the protonation states are sampled properly or barely move. They are specific to this kind of simulation and appear nowhere else. (UNVERIFIED) constant pH parameter A constant pH parameter giving how many simulation steps pass between attempts to change a residue's protonation state. Attempting more often samples the protonation states better and costs more, because each attempt requires extra calculation. Attempting too rarely leaves a residue stuck in one state for most of the run. (UNVERIFIED) CpHMD state change attempt frequency A constant pH parameter that specifies the number of solvent relaxation steps performed after a protonation state change. Those steps let the solvent reorient around the residue that has just been charged or neutralised. (UNVERIFIED) ntrelax CpHMD solvent relaxation steps number of CpHMD solvent relaxation steps ntrelax A simulation parameter controlling a run with a polarizable force field, covering how the induced dipoles are solved for and how tightly they must converge. A loose convergence makes the forces slightly inconsistent with the energy, which shows as energy drift. These settings do not arise at all with a fixed-charge force field. (UNVERIFIED) polarizable force field parameter A polarizable force field parameter giving how little the induced dipoles must change between iterations before they are accepted. A tighter value gives forces closer to the exact ones and costs more iterations each step. Too loose a value makes the forces slightly inconsistent with the energy, which shows up as energy drift. (UNVERIFIED) AMBER 26 Reference Manual, p. 465, entry dipole_scf_tol. induced dipole convergence criterion An input setting that determines the mathematical, statistical or filtering rules applied when existing simulation data is processed. Examples are the distance below which 2 structures count as similar, the size of a block in an error estimate, and the spacing of an analysis grid. The same trajectory analysed with different settings gives different numbers, so they are part of the result. (UNVERIFIED) analysis parameter A type of analysis parameter that controls a clustering algorithm used to group similar molecular conformations from a trajectory. These parameters define the distance metric, the cutoff for cluster membership, and the number of clusters to be generated. The clustering parameter settings are crucial for identifying the dominant conformational states sampled in a simulation. clustering parameter A clustering parameter that specifies the distance threshold for grouping conformations into the same cluster. Two structures with a distance (e.g., RMSD) less than the clustering distance cutoff are considered to be in the same cluster. This is a key parameter for defining the granularity of the clustering. This is often set with the keyword epsilon or cutoff. clustering distance cutoff A clustering parameter giving how many groups the algorithm should produce. Algorithms such as k-means require it in advance, so it is chosen either from prior knowledge or by trying several values and comparing a quality measure. Choosing it badly either merges distinct conformations or splits one into several. (UNVERIFIED) number of clusters to generate An analysis parameter that controls how a 3-dimensional grid of values is produced, such as a density or an electrostatic potential map. Its members set where the grid sits, how far apart its points are and what is computed at each. Spacing chosen too coarsely hides features; chosen too finely it produces files too large to handle. (UNVERIFIED) volumetric analysis parameter A volumetric analysis parameter giving the distance between neighbouring points of a 3-dimensional grid. A finer grid resolves smaller features and produces a file whose size grows with the cube of the refinement. Too coarse a grid can hide small features, such as the sites of ordered water molecules. (UNVERIFIED) AMBER 26 Reference Manual, p. 138, entry grdspc. grid spacing A volumetric analysis parameter naming which quantity is computed at each point of a 3-dimensional grid, such as a density, an interaction energy or a correlation function. The choice determines what the resulting map shows. The available choices depend on the analysis program. (UNVERIFIED) grid function A volumetric analysis parameter that specifies the number of points in each dimension of a three-dimensional grid. It controls the size and the resolution of the volumetric data. (UNVERIFIED) --ng grid points number of grid points --ng An analysis parameter controlling a statistical calculation, such as the confidence level, the number of resampling trials or the block size. These settings determine what an error bar means, so a value quoted without them cannot be interpreted. They are part of the result rather than an implementation detail. (UNVERIFIED) statistical analysis parameter A statistical analysis parameter giving the value below which a result is called significant. It is the probability of calling a result significant when there is in fact no effect, so setting it at 0.05 accepts such a false result in 1 of every 20 tests where no effect exists. It must be chosen before the data are examined, or it means nothing. (UNVERIFIED) p-value threshold A statistical analysis parameter giving the proportion of confidence intervals that would contain the true value if the whole procedure of sampling and calculating the interval were repeated many times. A level of 95% means 19 such intervals in 20 would contain it. It does not mean there is a 95% chance the true value lies in the particular interval computed. (UNVERIFIED) confidence level A statistical analysis parameter giving how many times the data are resampled in a bootstrap estimate. More resamples make the estimate of the uncertainty itself more stable, at a cost that grows in proportion. They do not make the underlying data more informative, which is a common misunderstanding of this setting. (UNVERIFIED) number of bootstrap replications A statistical analysis parameter giving the width of the smooth bumps placed on each data point in a kernel density estimate. Too narrow a width shows every random fluctuation as a peak; too wide a one merges genuinely separate peaks into one. It is the single most consequential choice in such an estimate. (UNVERIFIED) kernel density bandwidth An analysis parameter that controls how a free energy is calculated from simulation data. The group covers settings for end-point methods such as MM-PBSA, for thermodynamic integration, for the analysis of umbrella sampling and for 3D-RISM solvation calculations. The same simulation data analysed with different settings can give different free energies, so the settings are reported with the result. (UNVERIFIED) free energy analysis parameter A free energy analysis parameter that controls a molecular mechanics Poisson-Boltzmann or generalized Born surface area calculation. Its members set the continuum solvent model and the surface area treatment. Different choices for these settings are a common reason why published binding free energies for the same system disagree. (UNVERIFIED) MM-PBSA/GBSA parameter A dielectric constant that describes an implicit membrane slab in a continuum solvation calculation. It is set low to represent the non-polar lipid core, which screens charge about as poorly as a protein interior. It is what makes burying a charge in the membrane region cost energy in the model. (UNVERIFIED) AMBER 26 Reference Manual, p. 953, entry emem. membrane dielectric constant A parameter giving the radius of the sphere rolled over a molecule's van der Waals surface to define its solvent-accessible surface area. It is normally set to 1.4 ångström, the effective radius of a water molecule. It controls how far into clefts the solvent is considered to reach, and so it changes the non-polar part of the solvation free energy. (UNVERIFIED) AMBER 26 Reference Manual, p. 952, entry prbrad. probe radius A parameter giving the constant term added to the surface-area estimate of the non-polar solvation free energy. It accounts for the part of that energy which does not scale with area. It is fitted together with the surface tension coefficient, so the 2 should not be taken from different sources. (UNVERIFIED) cavity offset A parameter giving the coefficient that converts solvent-accessible surface area into non-polar solvation free energy. It has the same role as the surface tension of a large interface, but its value is obtained by fitting to data rather than by measurement. It is fitted together with the cavity offset. (UNVERIFIED) cavity surface tension A free energy analysis parameter that controls a thermodynamic integration calculation. Its members set how many intermediate coupling states are used and how they are spaced along the path. Too few or badly spaced states leave the integral under-sampled where the energy changes fastest. (UNVERIFIED) thermodynamic integration parameter A thermodynamic integration parameter giving how many intermediate coupling states are simulated along the transformation path. More windows make the numerical integration more accurate and the calculation proportionally more expensive. Too few leaves the integrand under-sampled where it changes fastest, which is usually near the end points. (UNVERIFIED) number of lambda windows A type of free energy analysis parameter that controls the analysis of umbrella sampling simulations to generate a potential of mean force. These parameters define the settings for the histogramming and reweighting procedure. The umbrella sampling analysis parameter settings are crucial for correctly combining the data from multiple windows. These are often set in the input file for a WHAM or MBAR analysis program. umbrella sampling analysis parameter An umbrella sampling analysis parameter that specifies the increment or step size used when assessing the convergence of the calculated Potential of Mean Force (PMF). The PMF is calculated using progressively larger amounts of data to check for convergence. The potential of mean force convergence step size is a key parameter for ensuring that the calculated free energy profile is reliable. This is a key parameter for WHAM analysis. potential of mean force convergence step size A parameter controlling a 3D-RISM calculation, covering the grid on which the solvent distribution is computed, the closure relation used, and how tightly the solution must converge. The closure choice changes the answer, not merely its precision. These settings must be reported for a solvation free energy from this method to be interpretable. (UNVERIFIED) 3D-RISM parameter A 3D-RISM parameter giving how little the solution must change between iterations before the calculation is declared finished. A tighter value gives a more accurate solvent distribution and needs more iterations. Too loose a value returns a result that looks complete and is still drifting. (UNVERIFIED) 3D-RISM convergence tolerance A 3D-RISM parameter giving the distance beyond which interactions between the solute and the solvent are not evaluated. Raising it improves accuracy and raises cost, because the number of grid points within that distance of each solute atom grows with the cube of the distance. It is one of the main controls on how long such a calculation takes. (UNVERIFIED) solute-solvent interaction cutoff A type of analysis parameter that controls the analysis of simulation data to generate properties that can be compared with Nuclear Magnetic Resonance (NMR) experiments. These parameters define the settings for calculating NMR-related observables from a trajectory. The NMR analysis parameter settings are crucial for validating simulations against experimental data. These are often set in the input file for an NMR analysis program. NMR analysis parameter An NMR analysis parameter that specifies a particular form of distance averaging used in NMR structure calculation and refinement. This method involves averaging the inverse sixth power of the distance between two atoms, which is directly related to the Nuclear Overhauser Effect (NOE) intensity. The <r⁻⁶>⁻¹/⁶ averaged distance is a key parameter for back-calculating NOEs from a simulation trajectory. This is often referred to as R6 averaging. R6 averaged distance A quantitative or qualitative output data item generated by a simulation or analysis process. It represents the actual measured state, derived metric, or calculated result of the system (e.g., instantaneous potential energy, calculated RMSD, assigned cluster IDs). (UNVERIFIED) computed observable A computed observable describing the thermodynamic state of a system, or the energy exchanged as it changes state, such as energy, temperature, pressure, entropy and free energy. Most of these depend only on the current state and not on how the system reached it, and are estimated as averages over sampled configurations. Work done in a nonequilibrium process is an exception, because its value depends on the path taken. (UNVERIFIED) thermodynamic property A thermodynamic property measuring a capacity to do work that belongs to a simulated system or to part of it, such as its total energy, a single term of the force field, or the interaction between 2 of its parts. In a classical simulation the potential part is computed from the atom positions using the force field, and the kinetic part from the velocities. The terms are reported separately at intervals through a run, because a fault usually shows in a single term rather than in the total. (UNVERIFIED) energy A component of a system's energy that depends only on where the atoms are, not on how fast they are moving. In a classical simulation it is computed from the force field, summing the bonded terms with the electrostatic and van der Waals interactions. The force on each atom is the negative of its gradient, and those forces drive the simulation forward. (UNVERIFIED) potential energy A potential energy summing every bonded and non-bonded contribution for 1 configuration of the system. It is reported at intervals during a run and is among the first quantities checked when a simulation is suspected of going wrong. It usually settles to a stable average once the system has equilibrated. (UNVERIFIED) total potential energy A component of the potential energy arising from the Coulomb interaction between the partial charges on the atoms. It falls off slowly with distance, so it reaches much further than the other non-bonded term and needs special treatment in a periodic system. It is what makes polarity, solvation and ionic binding come out right or wrong. (UNVERIFIED) electrostatic energy A component of the electrostatic energy covering only pairs of atoms separated by exactly 3 bonds. Such pairs are close enough that the plain Coulomb term would double-count what the torsion term already describes, so many force fields scale it down by a fixed factor. The factor differs between force fields, which is one reason their parameters cannot be mixed. (UNVERIFIED) 1-4 electrostatic energy A component of the potential energy covering the short-range interaction between atoms not bonded to each other: a steep repulsion when they approach too closely and a weaker attraction at slightly greater distance. The repulsion is what gives atoms an effective size. The attraction is what makes non-polar molecules condense into liquids at all. (UNVERIFIED) Van der Waals energy A component of the van der Waals energy covering only pairs of atoms separated by exactly 3 bonds. Such pairs are close enough that the full interaction would double-count what the torsion term already describes, so many force fields scale it down or use separate parameters for these pairs. The treatment differs between force fields, which is one reason their parameters cannot be mixed. (UNVERIFIED) 1-4 Van der Waals energy A component of the potential energy stored in the bending of the angle between 3 bonded atoms away from its preferred value. It is softer than the bond stretching term, so angles deform more readily than bond lengths. It is the term that keeps, for example, a carbon atom with 4 single bonds close to a tetrahedral shape. (UNVERIFIED) angle energy A component of the angle energy that adds a spring directly between the 2 outer atoms of a bonded triple, in addition to the angle term itself. It corrects the coupling between bending an angle and stretching the bonds around it, which the angle term alone cannot represent. It is used by the CHARMM force fields and not by most others. (UNVERIFIED) Urey-Bradley energy A component of the potential energy associated with rotation about a bond, defined by 4 atoms in sequence and repeating as the rotation goes round. It is what makes some rotational positions preferred over others, which sets the shapes a molecule can adopt. It is the term most often refitted when a force field is improved, because conformational preferences depend on it most. (UNVERIFIED) dihedral energy A component of the potential energy that holds a group of 4 atoms flat, or holds a centre in one of its 2 mirror forms. Unlike an ordinary torsion it is defined over atoms that are not in a chain, and in many force fields it is a harmonic term with a single preferred value rather than a repeating one. Without it, a planar group such as an aromatic ring or an amide could bend out of its plane. (UNVERIFIED) improper torsion energy A component of the potential energy arising from momentary fluctuations in an atom's electron cloud inducing matching fluctuations in its neighbours. It is always attractive and falls off rapidly with distance, so near neighbours contribute most. It is weak between any 2 atoms and dominant in large non-polar molecules simply because there are so many pairs. (UNVERIFIED) dispersion energy A component of the potential energy arising from a molecule's electron cloud being distorted by the electric field of its surroundings. A force field with fixed atomic charges cannot represent it and absorbs its average effect into those charges instead. Representing it explicitly is what distinguishes a polarizable force field and what makes it several times more expensive. (UNVERIFIED) polarization energy A component of the reported energy arising from artificial potentials added to hold atoms near chosen positions, or to keep a distance, angle or other value near a target. A large value shows that the system is being held far from where it would otherwise go. It is not part of the system's physical energy, so its effect must be removed or corrected for when thermodynamic quantities are computed. (UNVERIFIED) restraint energy A component of the reported energy computed quantum mechanically for one region of a system. It appears only in calculations where part of the system is treated that way and the rest with a force field. It cannot be compared with a force field energy, because the 2 are measured from different zero points. (UNVERIFIED) QM energy A QM energy giving the electronic energy of a region once the iterative self-consistent field (SCF) procedure has converged, as in a Hartree-Fock or density functional theory calculation. For Hartree-Fock it is the energy of the best single-configuration description of the electrons, which post-Hartree-Fock methods then improve by adding electron correlation. (UNVERIFIED) SCF energy A component of the system's energy that represents an artificial (non negative) energy term added to the potential to accelerate sampling in enhanced sampling simulations. This non-physical energy helps the system overcome high energy barriers more quickly. It is reported in the log files of methods like Accelerated MD as E_boost or amd_boost. boost energy Accelerated Molecular Dynamics (aMD or GaMD) is a powerful enhanced sampling technique designed to speed up conformational transitions in biomolecular simulations. It works by applying a harmonic boost potential to the system's true potential energy surface whenever it drops below a predefined threshold. The accelerated MD boost energy is the key quantity used to reweight the biased trajectory to recover correct thermodynamic properties. It is often reported in log files as amd_boost. accelerated MD boost energy A special potential term used in some older force fields to explicitly model hydrogen bonds. It is calculated using a 10-12 Lennard-Jones-like function between donor and acceptor atoms. This term is not commonly used in modern force fields but may be found in older simulation outputs. It is often reported as HBOND in energy summaries when such a potential is used. 10-12 H-bond energy A component of the reported energy contributed by the several copies of a region that a locally enhanced sampling simulation contains, which do not interact with each other. Because the force field parameters of the copies are scaled by 1 over the number of copies, the value is an average over the copies rather than the energy of a single real structure. The term exists only in this kind of simulation. (UNVERIFIED) locally enhanced sampling energy A thermodynamic property combining energy and entropy into the quantity that determines which way a process goes at fixed temperature. Its decrease in a process equals the largest amount of useful work the system can deliver under those conditions. Because it contains entropy it cannot be read off a single structure, only from an ensemble. (UNVERIFIED) free energy property A free energy giving the change on taking 2 molecules from separate, solvated states to a bound complex. It is the quantity that determines binding affinity, so it is what a docking or free energy calculation is usually trying to predict. It is negative when binding is favourable. (UNVERIFIED) binding free energy A free energy giving the difference in binding free energy between 2 ligands binding the same receptor. It is usually computed by transforming 1 ligand into the other alchemically, which cancels the parts of the calculation the 2 have in common. That cancellation is why it can be computed more accurately than either absolute value. (UNVERIFIED) relative binding free energy A free energy giving the change on moving a molecule from vacuum into a solvent. It measures how favourable it is for the molecule to be surrounded by that solvent, and so it underlies solubility and partitioning. It appears as a leg in most thermodynamic cycles used to compute binding. (UNVERIFIED) solvation free energy A solvation free energy where the solvent is water. It gives the change on moving a molecule from the gas phase into aqueous solution. It is the most common benchmark for testing whether a force field treats solute-water interactions correctly, because accurate experimental values exist for many small molecules. (UNVERIFIED) hydration free energy A component of the solvation free energy arising from electrostatic interaction between the solute's charges and the solvent, which polarizes in response to them. It is favourable for any charged or polar molecule moved from vacuum into a polar solvent, and the part of it lost when polar groups are buried, for example on binding, is called the desolvation penalty. Continuum solvent models compute this electrostatic part and the non-polar part separately. (UNVERIFIED) polar solvation energy A component of the solvation free energy arising from everything other than electrostatics, chiefly the cost of forming a cavity and the dispersion attraction between solute and solvent. It is often estimated as proportional to the solvent-accessible surface area, although some methods compute the cavity and dispersion parts separately. (UNVERIFIED) nonpolar solvation energy A non-polar solvation energy computed as a function of the solvent-accessible surface area of a molecule, usually a constant times the area plus an offset. It stands for the cost of opening a cavity in the solvent together with the dispersion attraction across its surface. It is the simplest of the non-polar treatments. (UNVERIFIED) surface area energy A polar solvation energy obtained by solving the Poisson-Boltzmann equation for a solute of low dielectric constant embedded in a continuum of high dielectric constant. It treats the electrostatics of solvation without approximating the shape of the molecule. It is more expensive and more faithful than the generalized Born alternative. (UNVERIFIED) Poisson-Boltzmann energy A polar solvation energy obtained from the generalized Born model, which approximates the Poisson-Boltzmann result with a closed-form expression using an effective radius for each atom. It is far faster, which is what makes it usable inside a running simulation rather than only in post-processing. Its accuracy depends on how the effective radii are computed. (UNVERIFIED) Generalized Born energy A specific type of solvation free energy calculated using the 3D Reference Interaction Site Model (3D-RISM). This method provides a detailed, atomistic picture of the solvent structure around the solute and its thermodynamics. The 3D-RISM solvation energy is a key output of 3D-RISM calculations. This value is often reported as the excess chemical potential. 3D-RISM solvation energy A type of thermodynamic property representing the free energy change of adding one particle to a system at constant volume and temperature. It is a direct measure of solvation free energy and is the primary output of methods like 3D-RISM. The calculated Excess Chemical Potential, often denoted µex or --exchem, is a key indicator of a molecule's solubility. This term is often found in the output of 3D-RISM calculations. excess chemical potential A thermodynamic property measuring how many microscopic arrangements a system can adopt while looking the same macroscopically. It enters the free energy with a negative sign multiplied by temperature, so a state with more arrangements is favoured, more strongly at higher temperature. It is far harder to compute from a simulation than the energy, because it depends on how much of the possible configurations the simulation visits, not only on an average over those it samples. (UNVERIFIED) entropy An entropy component giving the change in disorder when a solute is placed in a solvent. It combines the conformational freedom the solute loses with the reorganisation forced on the solvent around it. For a non-polar solute in water the solvent term dominates, which is the origin of the hydrophobic effect. (UNVERIFIED) solvation entropy An information entropy of the statistical weights given to configurations when data from several biased simulations are combined, measuring how evenly those weights are spread. A low value means a few configurations carry most of the weight, so the effective sample is much smaller than the number of frames. It is checked before a reweighted average, such as the free energy of a single bin of a free energy surface, is trusted. (UNVERIFIED) reweighting entropy An entropy component arising from the positions a molecule's centre of mass can occupy. It falls when a molecule is confined, which is why binding always carries a translational entropy cost. It depends on concentration, so the standard state must be stated with any value. (UNVERIFIED) translational entropy An entropy component arising from the orientations a molecule can adopt. Like the translational component it falls on binding, because a bound molecule is held in a fixed orientation. The 2 together are the rigid-body entropy cost of forming a complex. (UNVERIFIED) orientational entropy A thermodynamic property that should not change during a simulation in a given ensemble, and is therefore watched to check that the integration is sound. With no thermostat or barostat it is the total energy; with them it is that energy plus terms accounting for what they have added or removed. A steady drift in it means the timestep is too large or a setting is inconsistent, whatever else looks normal. (UNVERIFIED) conserved quantity A thermodynamic property giving the work done on a system while it is driven away from equilibrium, obtained by integrating the applied force along the path taken. It differs between repetitions of the same process, so it is a distribution rather than a single number. A fluctuation theorem relates that distribution to the equilibrium free energy difference. (UNVERIFIED) nonequilibrium work A type of output data descriptor that quantifies some aspect of the three-dimensional arrangement of atoms in a molecule or system. These descriptors are used to characterize the conformation, shape, size, and flexibility of molecules. They are the most common type of analysis performed on simulation trajectories. structural property A type of structural property that measures the average distance between the atoms of two superimposed molecular structures. It is the most common metric for quantifying the difference between two conformations, often used to track a simulation's stability relative to a starting structure. Analysis outputs frequently report the RMSD value over time. RMSD root mean square deviation (RMSD) A type of structural property that measures the fluctuation of each individual atom around its average position over the course of a simulation. It is used to identify flexible or mobile regions within a molecule, such as loops or termini in a protein. Analysis outputs often report the per-residue RMSF or atomicfluct to highlight these regions. RMSF root mean square fluctuation (RMSF) A type of structural property that measures the difference between two structures based on their internal pairwise atomic distances, rather than their Cartesian coordinates. This metric does not require structural alignment and can be more sensitive to changes in internal conformation. It is a useful alternative to standard RMSD for comparing flexible molecules. dRMSD distance RMSD (dRMSD) Hydrogen mass repartitioning is a system setup method where a portion of the mass of a heavy atom is transferred to its bonded hydrogen atoms. This makes the hydrogens heavier, which slows down their bond vibrations and allows a larger time step to be used in a molecular dynamics simulation. For experts, this technique, often abbreviated HMR, is a common strategy to increase computational efficiency without significantly altering the system's dynamics. hydrogen mass repartitioning analysis A distance between an ion and the oxygen atoms of the water molecules in its first solvation shell, usually taken as the position of the first peak of the ion-oxygen radial distribution function. It is a standard target when ion parameters are fitted or tested, because it can be compared directly with X-ray and neutron scattering data. (UNVERIFIED) ion-oxygen distance A structural property giving how far the line from a middle atom to a first atom is turned from the line from that middle atom to a third atom, for 3 specified atoms. Following it through a trajectory is the simplest way to watch a local geometry change, such as a hydrogen bond opening. It is measured between any 3 atoms, whether or not they are bonded. (UNVERIFIED) angle A type of structural property representing the torsion angle defined by four sequentially bonded atoms, which describes the rotation around the central bond. The dihedral angle is the most important descriptor for characterizing the overall conformation of a flexible molecule. Its value is a primary output of many structural analysis tools. This value is commonly referred to as dihedral or torsion. dihedral angle A dihedral angle describing the rotation about one of the bonds in a protein backbone. The pair of angles either side of each alpha carbon determines the local shape of the chain, and so which secondary structure that stretch forms. Plotting the 2 against each other gives the Ramachandran plot, whose crowded and empty regions show which shapes are possible at all. (UNVERIFIED) protein backbone torsion angle A specific type of dihedral angle that describes the conformation of an amino acid's side chain. These angles determine the spatial arrangement of the side chain's functional groups. They are critical for understanding molecular interactions and protein packing. These angles are commonly referred to as chi angles, such as chi1 or chi2. protein side-chain torsion angle A specific nucleic acid backbone dihedral angle is a torsion angle describing rotation about a bond in the sugar-phosphate backbone of DNA or RNA. The six backbone torsion angles (α, β, γ, δ, ε, and ζ), together with the glycosidic torsion angle (χ) and the sugar pucker, define the local conformation, flexibility, and overall geometry of the nucleic acid helix. nucleic acid torsion angle A type of structural property that measures the overall compactness of a molecule or a group of atoms. It is calculated as the root-mean-square distance of the atoms from their common center of mass. The radius of gyration is used to monitor large-scale conformational changes like protein folding or unfolding. Rg radius of gyration A structural property that quantifies the positional fluctuation or displacement of an atom, representing its mobility. It is calculated from the mean-square fluctuations of atomic positions in a simulation and is analogous to the crystallographic B-factor or temperature factor. High B-factor values indicate regions of high flexibility. This value is often referred to as bfactor or B-factor. B-factor A structural property rating how similar the overall folds of 2 proteins are, on a scale from 0 to 1, with residue distances weighted on a length-dependent scale so that the score does not depend on protein length. The weighting also makes it less sensitive than RMSD to a few badly matched regions. A value above roughly 0.5 means the 2 usually share the same fold. (UNVERIFIED) TM-score A structural property giving how many of the atom or residue contacts present in a reference folded structure are also present in the conformation being scored. It is usually reported as a fraction of the reference set rather than a raw count. It is the most common single-number reaction coordinate for folding. (UNVERIFIED) native contacts native contact A type of structural property that represents the average number of neighboring particles within a certain distance of a central particle. For an ion in water, the Coordination Number typically refers to the number of water molecules in its first solvation shell. It is a fundamental descriptor of local solvent structure and is used to validate force fields. This value is often abbreviated as CN. CN coordination number A type of structural property that measures the surface area of a molecule that is accessible to a solvent. This quantity is crucial for understanding solvation effects. It is a key component in implicit solvent models and is often used to characterize a molecule's exposure to its environment. This value is commonly known as Solvent Accessible Surface Area or SASA. molecular surface area A type of structural property that describes the specific three-dimensional conformation of a sugar ring, such as the ribose in DNA or RNA. The pucker is a critical determinant of the local and global structure of a nucleic acid helix. It is often described by a pseudorotation angle and can be analyzed to characterize A-form, B-form, or other helical types. This value is a standard output of nucleic acid analysis tools. puckering sugar pucker A structural property recording which of its 2 mirror forms a centre in a molecule takes. The 2 forms cannot be superimposed and often behave completely differently in a biological system. A force field cannot change one into the other, so an error here persists through an entire simulation unnoticed. (UNVERIFIED) chirality A type of structural property that describes the rotational position of a molecule or a group of atoms in three-dimensional space. It is often represented by a set of Euler angles, a rotation matrix, or a quaternion. Analyzing the orientation is essential for understanding the tumbling of molecules in solution or the alignment of components in a complex. orientation molecular orientation A specific type of orientation descriptor that represents the angle of rotation between the current orientation and a reference orientation. This provides a simple, scalar measure of how much a molecule has rotated. This value is often reported as ORIENTATION_ANGLE. orientation angle A specific type of orientation descriptor that represents the cosine of the angle of rotation between the current and reference positions. This value is often used in analyses of rotational correlation. This value is often reported as ORIENTATION_PROJ. orientation projection A base pair step parameter that describes the rotation of one base pair relative to the next around their short axes. A non-zero tilt causes the helix to bend. It is an important parameter for describing DNA curvature. This is a standard output of nucleic acid analysis programs. tilt A specific type of orientation descriptor that represents the angle of rotation of a molecule around a specific unit axis. The SPINANGLE is useful for analyzing spinning motions. spin angle A molecular orientation descriptor giving the rotation that best superimposes a chosen group of atoms onto a set of reference coordinates, written as a unit quaternion: 4 numbers that together encode a rotation angle and a rotation axis. Unlike Euler angles, a quaternion has no orientations at which the description breaks down, so it is a reliable way to follow how a molecule turns during a simulation. (UNVERIFIED) orientation quaternion A type of structural property that provides a detailed, quantitative description of the geometry of a nucleic acid helix. These parameters are calculated by specialized analysis programs to characterize the structure of DNA and RNA. They are essential for comparing simulated structures to experimental data and for understanding sequence-dependent structural variations. These are the standard output of programs like do_x3dna or cpptraj. nucleic acid parameter A nucleic acid parameter that describes the relative position and orientation of the 2 bases within a single base pair. It measures how far the pair departs from an ideal flat pairing, and includes shear, stretch, stagger, buckle, propeller twist and opening. (UNVERIFIED) base pair parameter A base pair parameter that quantifies the translational displacement of one base relative to the other along their short axis, within the plane of the base pair. A non-zero shear indicates a lateral displacement of the two bases from their ideal paired geometry, altering their relative alignment while maintaining the base-pair plane. This is a standard output of nucleic acid analysis programs. shear A base pair parameter that quantifies the translational displacement of one base relative to the other along the axis of the hydrogen bonds. A non-zero stretch measures how much the base pair is pulled apart or compressed. It reflects the strength and dynamics of the hydrogen bonds holding the pair together. This is a standard output of nucleic acid analysis programs. stretch A base pair parameter that quantifies the translational displacement of one base relative to the other along the axis perpendicular to the base pair plane. A non-zero stagger indicates that the two bases are offset in the direction normal to the base-pair plane, resulting in a difference in their relative height and a deviation from an ideal coplanar geometry. This is a standard output of nucleic acid analysis programs. stagger A base pair parameter that describes the rotation of the two bases relative to each other around their long axis, roughly parallel to the hydrogen bonds. A non-zero buckle angle indicates that the bases are tilted relative to each other, producing a bent base-pair geometry. It contributes to the overall shape of the helix. This is a standard output of nucleic acid analysis programs. buckle A base pair parameter that describes the rotation of the two bases relative to each other around an axis connecting their glycosidic bonds. This propeller twist optimizes stacking with adjacent base pairs but can strain hydrogen bonds. It is a key determinant of helix stability and groove dimensions. This is a standard output of nucleic acid analysis programs. propeller twist An intra-base-pair parameter that describes the relative rotation of the two bases around the axis perpendicular to the base-pair plane. A non-zero opening angle indicates that the base pair deviates from its ideal planar geometry, altering the orientation of the hydrogen-bonding and increasing their exposure to the solvent. Large opening values are often associated with transient base-pair opening ("breathing") or fraying events. Opening is a standard structural parameter reported by nucleic acid analysis programs. opening A type of nucleic acid parameter that describes the relative geometry and orientation between two successive base pairs in a helix. These parameters define the local helical structure. They are essential for characterizing the sequence-dependent variations in DNA and RNA structure. These parameters include shift, slide, rise, tilt, roll, and twist. base pair step parameter A base pair step parameter that describes the translational displacement of one base pair relative to the next along their short axes. A non-zero shift indicates a lateral sliding motion. It affects the alignment and stacking interactions between adjacent base pairs. This is a standard output of nucleic acid analysis programs. shift A base pair step parameter that describes the translational displacement of one base pair relative to the next along their long axes. A non-zero slide is a major determinant of the width of the major and minor grooves. It is a key parameter for describing B-form DNA. This is a standard output of nucleic acid analysis programs. slide A base pair step parameter that describes the translational distance between two adjacent base pairs along the central helical axis. The rise per step is a primary determinant of the overall length of the DNA or RNA helix. This is a standard output of nucleic acid analysis programs. rise A base pair step parameter that describes the rotation of one base pair relative to the next around their long axes. A non-zero roll angle causes the helix to bend into the major or minor groove. It is a crucial parameter for DNA bendability and protein recognition. This is a standard output of nucleic acid analysis programs. roll A base pair step parameter that describes the rotation of one base pair relative to the next around the central helical axis. The helical twist is the primary determinant of the number of base pairs per turn in a DNA or RNA helix. This is a standard output of nucleic acid analysis programs. twist A nucleic acid parameter that quantifies the position of a base pair's reference frame relative to the global or local helical axis. Displacement is the distance (in Å) between the center of the base pair and the global helical axis, measured perpendicular to the helix axis. They include x-displacement and y-displacement. displacement parameter A displacement parameter that measures the translation of a base pair away from the central helical axis towards the major or minor groove. The x-displacement is a key parameter for distinguishing between A-form and B-form DNA. This is a standard output of nucleic acid analysis programs. x-displacement A displacement parameter that measures the translation of a base pair away from the central helical axis along the axis of the base pair's hydrogen bonds. The y-displacement is typically small in canonical DNA structures. This is a standard output of nucleic acid analysis programs. y-displacement A computed observable describing how the electrons of a molecule are distributed and how they respond to a disturbance, usually obtained from a quantum mechanical calculation. Examples are atomic charges, dipole moments and polarizabilities. These properties govern chemical reactivity and how a molecule responds to light or to an electric field. (UNVERIFIED) electronic property An electronic property giving the partial electric charge assigned to a single atom in a molecule. No such quantity exists in the underlying physics, because electrons are not divided between atoms; it is a convention for splitting a continuous distribution. Different conventions give different values for the same molecule, so the method must always be stated. (UNVERIFIED) atomic charge An atomic charge obtained by fitting point charges on the atoms so that together they reproduce the electrostatic potential around the molecule computed by a quantum mechanical calculation. Such charges are widely used to build force field parameters for new molecules, because they reproduce how the molecule interacts electrically with its surroundings. Charges on buried atoms are poorly determined by the fit, which is why restrained variants such as RESP exist. (UNVERIFIED) electrostatic potential (ESP) derived charge An electronic property giving the size and direction of the separation between a molecule's positive and negative charge, equal to the charge multiplied by the distance separating it. It is what makes a molecule polar, and it determines how the molecule aligns in an electric field and how strongly it interacts with other polar molecules. For a neutral molecule it can be measured and does not depend on how charge is divided between atoms. (UNVERIFIED) dipole moment An electronic property describing a molecule's charge distribution in more detail than a dipole, through terms such as the quadrupole that capture progressively finer features of its shape. A molecule with no dipole can still interact strongly through these, as benzene does through its quadrupole. Some force fields, such as AMOEBA, place a dipole and a quadrupole on each atom as well as a point charge. (UNVERIFIED) multipole moment An electronic property giving how easily a molecule's electron cloud is distorted by an electric field. A larger, softer electron cloud distorts more, so polarizability rises with molecular size. It determines the strength of dispersion attraction, which is what holds non-polar molecules together. (UNVERIFIED) polarizability A polarizability describing how a molecule's electron cloud responds to a strong electric field, beyond the part that responds in proportion to the field. It is what makes a material double the frequency of light passing through it, or change its refractive index with intensity. It matters for optical materials and is negligible at ordinary field strengths. (UNVERIFIED) hyperpolarizability A polarizability describing how a molecule's response to an oscillating electric field, such as light, depends on the frequency of the oscillation. The response grows as the frequency approaches that of an electronic excitation of the molecule, so the same calculation also gives the energies at which the molecule absorbs light. It determines how the refractive index changes with the colour of light. (UNVERIFIED) time-dependent polarizability An electronic property giving the least energy needed to remove an electron from a molecule. A high value means the molecule holds its electrons tightly and is hard to oxidise. Together with the electron affinity it says how readily the molecule will give or take electrons in a reaction. (UNVERIFIED) ionization potential An electronic property giving the energy released when an electron is added to a molecule. A large value means the molecule accepts electrons readily and is easily reduced. It is the counterpart of the ionization potential, which measures losing one instead. (UNVERIFIED) electron affinity An electronic property read directly from the mathematical function describing a molecule's electrons, rather than from an energy difference. Orbital energies, orbital shapes and the electron density itself belong here. They depend on the method and basis set used, so they are comparable only between calculations done the same way. (UNVERIFIED) wavefunction property A wavefunction property giving the total electronic energy of a system at a fixed arrangement of its nuclei. It is the main output of a quantum chemistry calculation and the quantity whose derivatives give the forces. It is meaningful only as a difference, because its absolute value depends on the method and basis set. (UNVERIFIED) wavefunction energy A wavefunction property saying whether a converged self-consistent field solution is a true energy minimum or only a saddle point, from which a small change in the orbitals would lower the energy. Such a solution passes the convergence test even though a lower-energy solution exists. Checking this is routine for open-shell and transition metal systems, where such unstable solutions are common. (UNVERIFIED) wavefunction stability index An electronic property giving the total charge of a chosen group of atoms rather than of a single one. Grouping is more robust than per-atom charges, because the arbitrariness of dividing charge between neighbouring atoms partly cancels. It is used to state the charge of a quantum region or of a functional group. (UNVERIFIED) charge distribution descriptor A charge distribution descriptor giving the net charge of the core region in an adaptive QM/MM calculation, the atoms that are treated with quantum mechanics throughout the simulation. It is given as a whole number, because the quantum calculation needs the total charge of the atoms it treats. (UNVERIFIED) core region charge A charge distribution descriptor giving the net charge of the buffer region in an adaptive buffered force-mixing QM/MM calculation, the shell of atoms around the atoms that move with quantum mechanical forces. Buffer atoms are included in the larger quantum calculation so that the atoms inside get accurate forces, but their own forces from that calculation are discarded. Their net charge is given as a whole number, because the quantum calculation needs the total charge of the atoms it treats. (UNVERIFIED) buffer region charge A computed observable describing how a system moves and changes with time, rather than what shape it has at one instant. Diffusion coefficients, relaxation times and transition rates belong here. They are normally computed from a trajectory, because a single structure carries no information about time. (UNVERIFIED) kinetic property A kinetic property giving how fast particles spread out through random thermal motion. It is obtained from the slope of the mean squared displacement against time, or from the integral of the velocity autocorrelation function. It is one of the few simulated quantities that can be compared directly with a straightforward experiment. (UNVERIFIED) diffusion coefficient A kinetic property giving the average of the squared distance a particle has moved from where it started, as a function of elapsed time. In a liquid it grows in proportion to time once the motion has become random, and the constant of proportionality gives the diffusion coefficient. Departure from that straight line marks motion that is still constrained rather than free. (UNVERIFIED) mean square displacement A type of kinetic property that characterizes the rate at which nuclear spins return to thermal equilibrium after being perturbed in a Nuclear Magnetic Resonance (NMR) experiment. These relaxation times are sensitive to molecular motions on different timescales. They are key experimental observables that can be compared with simulation results to validate molecular dynamics. These are known as T1 and T2 relaxation times. NMR relaxation time A specific type of NMR relaxation time, also known as the spin-lattice or longitudinal relaxation time. It characterizes the rate at which the nuclear spins exchange energy with their surrounding environment (the "lattice"). The T1 relaxation time is sensitive to fast molecular motions, on the order of the inverse of the Larmor frequency. This is a key parameter in NMR spectroscopy. T1 relaxation A specific type of NMR relaxation time, also known as the spin-spin or transverse relaxation time. It characterizes the rate at which the nuclear spins lose phase coherence with each other. The T2 relaxation time is sensitive to both fast and slow molecular motions, making it a rich source of dynamical information. This is a key parameter in NMR spectroscopy. T2 relaxation A type of kinetic property that represents the time scale over which a system's properties, such as the orientation of a bond, remain correlated. It is a measure of the "memory" of the system's dynamics. The correlation time is a key parameter in the interpretation of NMR relaxation data. This is often denoted tcorr or tau_c. correlation time A physical quantity that can be measured by or is related to a spectroscopic experiment. These properties are often calculated from simulations and compared directly with experimental data to validate the accuracy of the simulation model. They provide a crucial link between computational models and the real world. These are the primary outputs of computational spectroscopy. spectroscopic property A type of spectroscopic property that measures the through-bond interaction between two nuclear spins, which causes splitting of NMR signals. The J-coupling is highly sensitive to the dihedral angle between the coupled nuclei, as described by the Karplus relationship. It is a key piece of experimental data used to determine molecular conformation. This is a standard output of NMR analysis programs. J-coupling A type of spectroscopic property that provides long-range structural information in NMR experiments by measuring the orientation of bond vectors in a weakly aligned molecule. Residual Dipolar Couplings (RDCs) are extremely valuable for determining the relative orientation of distant parts of a molecule. They provide powerful restraints for refining molecular structures. residual dipolar coupling The deuterium order parameter (SCD) measures the orientational order and dynamic disorder of lipid acyl chains in MD simulations. It allows for direct validation of membrane fluidity and structural phases against deuterium NMR experiments. deuterium order parameter A type of spectroscopic property that measures the motional restriction of a bond vector, typically an N-H bond in a protein, on fast timescales. The S2 order parameter ranges from 0 (completely isotropic motion) to 1 (completely rigid) and is measured by NMR relaxation experiments. It is a powerful tool for characterizing internal protein dynamics. This is a standard output of NMR relaxation analysis. S2 order parameter A computed observable obtained by applying a statistical measure to a dataset, such as a set of structures from a trajectory or a set of predicted and reference values. Examples are error measures, distances and divergences between data sets or distributions, and clustering quality scores. (UNVERIFIED) statistical property OBSOLETE. A specific type of correlation coefficient that measures the linear relationship between two continuous variables. It is the most common method for calculating correlation. This value is often reported in analysis as the Pearson correlation coefficient or simply 'r'. This is a standard output of statistical packages. Obsoleted 2026-07-29: duplicated both MOLSIM:000112 and STATO:0000280 - 'Pearson product-moment correlation coefficient' is the full formal name of Pearson's correlation coefficient. Replaced by the reused STATO term. obsolete Pearson product-moment correlation coefficient true OBSOLETE. A specific type of correlation coefficient that measures the monotonic relationship between two variables, based on their ranks rather than their raw values. It is less sensitive to outliers than the Pearson correlation. The Spearman rank correlation coefficient is a non-parametric measure of association. This is often reported as rho or Spearman's rho. Obsoleted 2026-07-29: duplicated STATO:0000201. STATO is the canonical OBO home for statistics and its definition of this term is fuller than MOLSIM's (it covers the monotonic-function reading, the perfect +/-1 case, and when to prefer it over Pearson's). Note the symbol 'rho' named in the retired definition is not carried on the STATO term, which has no synonyms, because MOLSIM must not annotate a foreign IRI. obsolete Spearman rank correlation coefficient true A correlation coefficient that measures how far 2 variables place items in the same order, computed by counting the pairs of items they order the same way and the pairs they order differently. It uses only the order of the values, so it does not change when the values are rescaled and it is much less sensitive to extreme outliers than a coefficient based on the values themselves. It suits comparisons where the ranking matters more than the exact numbers. (UNVERIFIED) Kendall tau correlation coefficient A statistical property giving the typical size of the difference between predicted and observed values, computed by squaring the differences, averaging them and taking the square root. Squaring first makes large errors count disproportionately, so a single bad prediction moves it noticeably. It is in the same units as the quantity itself, which makes it directly interpretable. (UNVERIFIED) root mean square error (RMSE) A statistical property giving the average size of the difference between predicted and observed values, ignoring sign. Because it does not square the differences, every error counts in proportion to its size and one large error does not dominate. It is the measure to prefer when outliers are expected and should not be allowed to control the score. (UNVERIFIED) mean absolute error (MAE) OBSOLETE. A type of statistical property that estimates the standard deviation of the sample mean, quantifying how precisely the sample mean represents the true population mean. It is calculated by dividing the sample standard deviation by the square root of the sample size. The standard error of the mean is a key measure of the uncertainty of a calculated average value. This is often abbreviated as SEM. Obsoleted 2026-07-29: duplicated STATO:0000037. STATO is the canonical OBO home for statistics; reused in place of this term. Note the '(SEM)' abbreviation carried by the retired label is not reproduced on the STATO term, because MOLSIM must not annotate a foreign IRI; STATO's own definition spells out the abbreviation. obsolete standard error of the mean (SEM) true OBSOLETE. A type of statistical property, also known as R-squared, that indicates the proportion of the variance in the dependent variable that is predictable from the independent variable(s). It provides a measure of how well the observed outcomes are replicated by the model. The coefficient of determination is a key metric for assessing the goodness of fit of a regression model. This is often reported as R^2 or R-squared. Obsoleted 2026-07-29: duplicated STATO:0000564. STATO is the canonical OBO home for statistics; reused in place of this term. obsolete coefficient of determination true A type of statistical property that quantifies the degree to which two probability distributions share common regions. In umbrella sampling simulations, this is a critical diagnostic for ensuring that adjacent simulation windows have sampled sufficiently overlapping regions of phase space. The distribution overlap is a key metric for validating the results of umbrella sampling or replica exchange simulations. This is often reported as Phase space overlap (PSO). distribution overlap A measure of how far the configurations sampled at 2 neighbouring alchemical states coincide. Free energy estimators need configurations typical of 1 state to be reachable from the other; where they are not, the estimate is biased whatever its apparent precision. It is checked per pair of neighbouring states, not once for the whole calculation. (UNVERIFIED) phase space overlap A statistical property giving how likely something is, given that something else is already known to have happened. The transition probabilities of a Markov model are of this kind: the chance of being in 1 state next, given the state now. (UNVERIFIED) conditional probability OBSOLETE. A type of statistical property that measures the amount of variation or dispersion of a set of values. A low standard deviation indicates that the values tend to be close to the mean, while a high value indicates that the values are spread out over a wider range. It is a fundamental measure of the variability of a calculated property. This is often abbreviated as sd or stdev. Obsoleted 2026-07-29: duplicated STATO:0000237. STATO is the canonical OBO home for statistics; reused in place of this term. obsolete standard deviation true A statistical property giving how far apart 2 sets of points are, as the largest distance from any point of either set to the nearest point of the other. It is controlled by the worst-matched point rather than by the typical one. That makes it sensitive to a single outlying point, which is useful when the worst match matters and misleading when it does not. (UNVERIFIED) Hausdorff distance A statistical property rating how well a set of groups describes the data, typically by comparing how tight the groups are against how far apart they lie. Such measures are used to choose how many groups to make, since most algorithms will produce any number asked for. No measure settles the question; they narrow it. (UNVERIFIED) clustering quality metric A clustering quality metric giving the average, over all groups, of how similar each group is to the one it most resembles. Similarity here combines how spread out the 2 groups are with how close their centres lie. A lower value means the groups are compact and well separated, so it is minimised rather than maximised. (UNVERIFIED) Davies-Bouldin Index A clustering quality metric giving the ratio of the variation between groups to the variation within them. A high value means the groups differ from each other far more than their members differ among themselves. It is computed for several possible numbers of groups, and the number giving the highest value is taken as the best. (UNVERIFIED) pseudo-F statistic A clustering quality metric giving the share of the total variation in the data that is accounted for by the differences between groups. A value near 1 means almost all the variation is captured by the grouping. It always rises as more groups are added, so it is read by looking for where it stops rising steeply rather than by taking its maximum. (UNVERIFIED) SSR/SST ratio A clustering quality metric giving, for each structure, how much closer it is to its own group than to the nearest other group, on a scale from −1 to 1. A value near 1 means it clearly belongs where it was put, near 0 that it sits on a boundary, and negative that it would fit another group better. Averaging it over all structures rates the whole grouping. (UNVERIFIED) cluster silhouette A statistical property giving how strongly 2 quantities are related when separated by a given distance or a given interval of time. Its shape says both how strong the relationship is and over what range it persists. Many quantities taken from a trajectory, such as radial distribution functions and diffusion coefficients, are correlation functions or are derived from them. (UNVERIFIED) correlation function A correlation function giving how much a particle's velocity at one moment still resembles its velocity a given time later. It starts at its maximum and decays as collisions with neighbouring particles change the direction of motion. Its integral over time is proportional to the diffusion coefficient, a Green-Kubo relation. (UNVERIFIED) VACF velocity autocorrelation function A specific type of correlation function that measures the covariance of a signal with itself at different time points. It is similar to the autocorrelation function but is not normalized. auto-covariance A correlation function giving the total departure from randomness in the positions of 2 particles in a liquid, counting both their direct influence on each other and influence passed through intervening particles. It is written h(r) and relates to the pair distribution function as h(r) = g(r) − 1. It is 1 of the 2 unknowns the RISM equations solve for. (UNVERIFIED) total correlation function A specific type of correlation function that describes the direct correlation between two particles in a liquid, excluding indirect effects mediated by other particles. It is a fundamental quantity in the statistical mechanics of liquids, related to the total correlation function by the Ornstein-Zernike equation. This is often denoted as c(r). (UNVERIFIED) direct correlation function A correlation function giving how likely a particle is to be found at each distance from another particle, relative to a uniform distribution. Peaks mark the distances at which neighbours prefer to sit: the first peak gives the most likely distance of the nearest neighbours, and the first minimum after it marks the outer edge of the first solvation shell. It is the standard way the structure of a liquid is described, and it can be compared with X-ray and neutron scattering data. (UNVERIFIED) pair-distribution function A type of output data descriptor for calculated properties that do not fit into the other major categories like thermodynamic or structural properties. This includes mathematical quantities derived from the potential energy surface or other specific analysis metrics. These are often specialized outputs from analysis programs. miscellaneous descriptor A vector of the first derivatives of the potential energy with respect to each atomic coordinate. Its negative is the force on each atom, which is what a simulation uses to move them. A minimization has converged when it is close to zero everywhere. (UNVERIFIED) gradient A matrix of the second derivatives of the potential energy with respect to every pair of atomic coordinates. It gives the curvature of the energy surface, so at a point where the gradient is zero it shows whether that point is a minimum, a saddle point or a maximum. Diagonalising it after weighting by the atomic masses gives the vibrational normal modes and their frequencies. (UNVERIFIED) hessian A descriptor giving how many interactions of a chosen kind, such as hydrogen bonds or close contacts, involve a particular atom or group. It reduces a pattern of contacts to a single number that can be followed over time. What it counts depends entirely on the distance and angle criteria used, which must be stated with it. (UNVERIFIED) atom interaction count A descriptor giving a single number that expresses how alike 2 entities are, such as 2 molecular structures or 2 atom types, according to a stated measure. The scale and range depend on the measure chosen, such as the Tanimoto coefficient for molecular fingerprints. Such scores are used, for example, to rank compounds in virtual screening. (UNVERIFIED) similarity score The Adaptive String Method (ASM) is a numerical technique used in computational chemistry and physics to find minimum free energy paths (MFEPs) or reaction pathways in complex, high-dimensional systems such as chemical reactions, conformational changes in biomolecules, or phase transitions. It is an improvement over the original string method because it adapts the distribution of points ("images") along the path to better capture regions of high curvature or complexity. For experts, this adaptive placement of images leads to a more efficient and accurate determination of the minimum free energy path. adaptive string method Ensemble trajectory analysis is a method that involves analyzing a large collection of independent simulation trajectories together, rather than just a single long one. This approach provides a more robust and statistically reliable picture of a molecule's behavior by sampling a wider range of starting conditions and potential pathways. It is particularly important for studying complex processes like protein folding or for achieving converged statistics on a system's properties. ensemble trajectory analysis A ligand charge model that keeps the AM1-BCC approach, charges from a semi-empirical AM1 calculation followed by bond charge corrections, but with the corrections refitted for use with the second general AMBER force field (GAFF2). The corrections were fitted so that GAFF2 with these charges reproduces measured solvation free energies, such as hydration free energies, of organic molecules. AMBER recommends combining it with GAFF2 rather than with the first general AMBER force field. (UNVERIFIED) ABCG2 charge model A method that obtains atomic charges by taking the charges from a fast semi-empirical calculation and correcting them with terms fitted to reproduce measured dipole moments. The correction is what fixes the known bias of the underlying charges. It is fast enough to apply to large numbers of molecules. (UNVERIFIED) CM1 charge method A charge model refining the earlier CM1 scheme with corrections fitted against a larger set of measured dipole moments and covering more elements. The functional form is unchanged; the fitting set and the coverage are wider. It is used where semi-empirical charges are wanted quickly and reliably. (UNVERIFIED) CM2 charge model A charge model that assigns partial charges by passing charge between bonded atoms according to their electronegativity until the flow stops. It needs no quantum calculation at all, only the bond connectivity. It is fast enough for millions of molecules and crude enough that it is used for screening rather than for simulation. (UNVERIFIED) Gasteiger charge model A charge model in which atomic charges are recomputed in response to the surrounding electric field, and the calculation repeated until they stop changing. Each atom's charge affects the field its neighbours feel, so they cannot be solved one at a time. It is one of the ways a polarizable force field represents the response of the electron cloud. (UNVERIFIED) iterative polarizable charge model A force field for RNA, built on AMBER ff99 with the bsc0 alpha and gamma backbone correction and with the glycosidic torsion chi refit by the Olomouc group. The refit removed the shift of chi into the high-anti region that made RNA duplexes open into non-native ladder-like structures in long simulations. It is the RNA force field recommended in AMBER. (UNVERIFIED) AMBER OL3 This is a specialized force field for simulating proteins that pays extra attention to how proteins interact with water and charged ions. The AMBER ff15ipq force field was developed to improve the description of protein interactions with ions by using a novel charge model (Implicitly Polarized Charge, IPq) that better reproduces experimental hydration free energies. This force field is particularly recommended for simulations where the accurate modeling of protein-ion interactions or salt bridge dynamics is critical to the system's behavior. AMBER ff15ipq A force field specifically parameterized for the simulation of protein mimetics, which are synthetic molecules designed to mimic natural proteins. It extends the existing ff15ipq force field to include parameters for non-standard, artificial, heterogenous building blocks, such as D-amino acids and beta-amino acids. This enables researchers to accurately model the structure and dynamics of these non-natural biopolymers. AMBER ff15ipq-m Amber ff03 is a modified version of the ff99 AMBER force field. The main changes are that charges are now derived from quantum calculations that use a continuum dielectric to mimic solvent polarization, and that the φ and ψ backbone torsions for proteins are modified, with the effect of decreasing the preference for helical configurations. The changes are just for proteins; nucleic acid parameters are the same as in ff99. AMBER ff03 This is a simplified version of the AMBER ff03 force field where the aliphatic hydrogen atoms on all amino acid side-chains are united to their corresponding carbon atoms. The aliphatic hydrogen atoms on all alpha carbon atoms are still represented explicitly to minimize the impact of the united-atom approximation on protein backbone conformations. In addition, aromatic hydrogens are also explicitly represented. Nucleic acid parameters are still in all atom and kept the same as in ff99. AMBER ff03ua A force field for DNA, built on AMBER ff99 with the bsc0 correction, that adds 3 refinements from the Olomouc group: to the epsilon and zeta backbone torsions, to the glycosidic torsion chi, and to the beta backbone torsion. It improves the description of the sugar-phosphate backbone and of helical twist over the preceding DNA parameters. It was later extended by OL21 and then by OL24, which AMBER now recommends for DNA. (UNVERIFIED) AMBER OL15 A nucleic acid force field is a set of mathematical functions and parameters specifically designed to simulate the behavior of DNA and RNA molecules. These force fields contain parameters optimized to accurately reproduce the unique structural features of nucleic acids, such as the double helix, sugar puckering, and base stacking interactions. They are essential for any realistic simulation of genetic material or functional RNA molecules. nucleic acid force field The AMBER bsc1 force field is a specific and widely used refinement for the dihedral angle parameters of the DNA backbone, designed for use with AMBER force fields. Developed by the Barcelona Supercomputing Center (BSC), it significantly improves the description of DNA's structural dynamics compared to the uncorrected force fields. It is considered a standard addition for obtaining accurate DNA simulations. AMBER bsc1 A force field for DNA whose bonded parameters and partial charges were derived from high-level quantum mechanical calculations rather than fitted to experimental data. The consistent derivation is what distinguishes it from force fields assembled from separately fitted terms. It targets DNA structure and dynamics specifically. (UNVERIFIED) tumuc1 DNA force field A force field for DNA from the Olomouc group that adds a refinement of the alpha and gamma backbone torsions to OL15. It was followed by OL24, which adds a refinement of the sugar pucker and is now the DNA force field AMBER recommends. (UNVERIFIED) AMBER OL21 A force field for phospholipids in the AMBER family, introduced in 2012, that provided the first modular framework for lipids in AMBER, in which head groups and tails are parameterised separately and combined. The modularity allows lipids not in the original set to be assembled from existing pieces. It has been superseded by later revisions in the same series. (UNVERIFIED) lipid11 A lipid force field built on the general AMBER force field, with its non-bonded parameters and the related torsion terms refined specifically for lipids. It was the first lipid parameter set based on the AMBER energy function that allowed lipid bilayers to be simulated at constant pressure without an applied surface tension. It was introduced in 2012, together with the modular Lipid11 framework. (UNVERIFIED) GAFFLipid A classical force field parameter set fitted to reproduce the geometries of proteins energy-minimized with the AM1 semi-empirical quantum method. It is used to pre-minimize a protein cheaply before a more expensive AM1 calculation, so that the starting geometry is already close to the AM1 one. It was meant only to reproduce AM1 geometries and was not tested for other uses, such as classical molecular dynamics. (UNVERIFIED) ParmAM1 A classical force field parameter set fitted to reproduce the geometries of proteins energy-minimized with the PM3 semi-empirical quantum method. It is used to pre-minimize a protein cheaply before a more expensive PM3 calculation, so that the starting geometry is already close to the PM3 one. It is the PM3 counterpart of ParmAM1, and was not tested for other uses, such as classical molecular dynamics. (UNVERIFIED) ParmPM3 A force field parameterised for solid-state materials such as metals, oxides, semiconductors and polymers, or for their interfaces with liquids. It is fitted to reproduce material properties such as crystal structure, elastic constants and defect energies, rather than the behaviour of biomolecules in water. Many such force fields use many-body terms, because bonding in a metal or a semiconductor cannot be described by fixed pairs of bonded atoms. (UNVERIFIED) material force field A material force field that describes the interaction between water and the (111) crystal face of platinum, developed as part of a project on multiscale simulations in catalysis. It adds Lennard-Jones, Gaussian and angle-dependent terms between surface platinum atoms and water oxygens, and can be combined with any water model. At present it has parameters only for water over platinum (111). (UNVERIFIED) GAL17 force field A force field parameter set for the fluorescent dye molecules used to label biomolecules in experiments. It is built to be used alongside the standard AMBER protein and nucleic acid parameters without modification. It lets a labelled construct be simulated as it was actually measured, which matters when interpreting FRET distances. (UNVERIFIED) AMBER-DYES force field A semi-empirical quantum method applying pairwise distance-dependent corrections to the PM3 parameterisation. The corrections markedly improve heats of formation for organic molecules. It uses the same underlying approximation and differs in the correction terms and their fitted values. (UNVERIFIED) PDDG/PM3 A semi-empirical quantum method extending MNDO with d orbitals on the heavier main-group elements. Without them, elements from the third row onwards are described badly, because their bonding involves those orbitals. It extends the method's usable range rather than changing how it treats the elements it already covered. (UNVERIFIED) MNDO/d A model that adds the effect of a field imposed from outside the system, such as an electric or magnetic field. The field acts on the charges in the system but is not affected by them. It is how a simulation represents an applied voltage, or the environment of a membrane under a potential difference. (UNVERIFIED) external field model An external field model in which the field comes from one or more fixed charges placed in or around the system. The charges do not move and feel no force themselves. It is a way of representing a nearby charged object without simulating it. (UNVERIFIED) external point charge model The simplest approximation used for the exchange-correlation term in density functional theory, which treats the electron density at each point as though it were part of a uniform gas of that density. It uses no information about how the density varies from place to place. It is reasonable for metals and poor for molecules, where the density changes sharply. (UNVERIFIED) LDA local density approximation A class of approximations for the exchange-correlation energy in density functional theory that uses both the electron density at each point and its gradient, that is, how fast the density changes there. Including the gradient corrects much of the error that the local density approximation makes for molecules. PBE and BLYP are widely used examples. (UNVERIFIED) GGA generalized gradient approximation An exchange-correlation functional that mixes a fixed fraction of exact exchange, computed as in Hartree-Fock theory, with exchange and correlation from a generalized gradient approximation (GGA) functional. The mixture cancels part of the error each part makes alone, particularly the artificial repulsion of an electron from itself. It costs more than a pure GGA functional because exact exchange is expensive to compute, and B3LYP is the best-known example. (UNVERIFIED) hybrid GGA functional A constituent part of a hybrid model in which a small region is treated with quantum mechanics and the rest with a force field. The parts include the choice of quantum and classical regions, the quantum method and its settings, such as the basis set and integration grid, corrections such as an empirical dispersion term, and the treatment of the boundary between the regions. The boundary is the difficult part, because bonds crossing it must be handled without either description failing. (UNVERIFIED) QM/MM model component A set of points at which a function is evaluated so that an integral over space can be computed as a weighted sum. Density functional theory needs one because part of its energy cannot be integrated analytically. Its density controls both the accuracy of the result and a noticeable share of the cost. (UNVERIFIED) integration grid A standard pruned integration grid for density functional theory calculations, proposed by Gill, Johnson and Pople in 1993, with 50 radial points and up to 194 angular points around each atom. Pruning means fewer angular points are used near the nucleus and far from it, where the integrand varies less, which keeps the grid small. It is a small, low-cost grid accurate enough for many routine calculations, and because it is defined in advance, every program that implements it uses the same grid. (UNVERIFIED) SG1 integration grid An extra atom, usually a hydrogen, placed in a hybrid quantum and classical simulation to complete a bond cut by the boundary between the 2 regions. Without it the quantum region would have an unsatisfied valence and behave like a radical. It is not part of the real molecule and is invisible to the classical region. (UNVERIFIED) link atom The part of a force field giving how the energy rises as a covalent bond departs from its preferred length. It is almost always a simple spring, chosen because it is cheap and accurate for the small departures thermal motion produces. A spring cannot break, which is the deliberate limitation of a classical force field. (UNVERIFIED) bond stretching potential A force field component that gives how the energy rises as the angle between 3 bonded atoms moves away from its preferred value. It is usually a harmonic term, which behaves like a spring much softer than the one for bond stretching. Together with the bond term, it sets the local geometry of a molecule. (UNVERIFIED) angle bending potential A force field component that gives how the energy varies with rotation about a bond, measured by the dihedral angle of 4 consecutive bonded atoms. It is usually written as a sum of cosine terms, so it repeats through a full turn, and the terms included set how many minima there are per turn. It is the main term that sets which shapes a flexible molecule prefers. (UNVERIFIED) dihedral potential It is an electrostatic or van der Waals force between two atoms that are separated by at least three covalent bonds. It is typically composed of two parts: the van der Waals component, which models short-range repulsion and dispersion, and the electrostatic component, which models the Coulomb interaction between atomic partial charges. 1-4 non-bonded interaction A point in a force field that carries charge or interaction parameters but no mass, held at a fixed position relative to real atoms. It lets charge be placed where no nucleus sits, such as on a lone pair or slightly off a nucleus. The 4-point water models and the extra points on halogens are the common examples. (UNVERIFIED) extra points virtual site A way of modelling a metal ion with explicit bonds to the atoms coordinating it, so the geometry of the site is held by the same terms that hold any other bond. It reproduces the fixed arrangement of a metal centre in a protein. The bonds cannot break, so the ion cannot exchange its coordinating atoms or leave. (UNVERIFIED) bonded metal model A way of modelling a metal ion as a charged sphere interacting only through electrostatics and van der Waals terms, with no bonds to its neighbours. The ion can then exchange its surroundings freely, as a real ion in solution does. It cannot reproduce the fixed geometry of a metal held in a protein site. (UNVERIFIED) non-bonded metal model A non-bonded potential that adds a term falling off with the fourth power of distance to the usual repulsion and attraction, representing the ion polarizing the molecules around it. That extra attraction is what a fixed-charge model otherwise omits entirely for a highly charged ion. It lets one parameter set reproduce both an ion's hydration energy and its ion-water distance, which the standard form cannot. (UNVERIFIED) 12-6-4 Lennard-Jones model A solvation model in which the solute, or the part of it being studied, is surrounded by a spherical droplet of explicit water rather than a fully periodic box. A restraining potential at the edge of the sphere usually keeps the water molecules from leaving. Using a droplet reduces the number of water molecules sharply, but water near the boundary behaves unlike bulk water, so results there are less reliable. (UNVERIFIED) water cap A representation of a single molecule as a set of atoms together with the rules governing how they interact. It fixes what is represented explicitly and what is left out, such as whether hydrogens appear individually. It is the unit a simulated system is assembled from. (UNVERIFIED) molecule model A model in which a molecule is represented as a set of independent harmonic oscillators, each vibrating about a fixed position. It is not meant to describe any real molecule; it serves as a reference state whose free energy is known exactly. That known value is what absolute free energies of solids and of bound states are measured against. (UNVERIFIED) Einstein molecule model An implicit solvent model that approximates the electrostatic part of the solvation free energy with a closed-form expression, using an effective radius for each atom that reflects how deeply it is buried. It avoids solving a partial differential equation, so it is fast enough to run inside a simulation rather than only afterwards. Its accuracy rests entirely on how those effective radii are computed. (UNVERIFIED) generalized Born implicit solvent model A generalized Born implicit solvent model that keeps the functional form of the GBn neck model but refits its parameters, with separate values for each element and for proteins and nucleic acids. It reproduces Poisson-Boltzmann and explicit solvent results more closely than the earlier neck model. AMBER recommends it, combined with the ff14SBonlysc force field, for proteins, nucleic acids and their complexes. (UNVERIFIED) GB-Neck2 implicit solvent model An experimental refinement model is a computational model used to improve or interpret raw experimental data, particularly in structural biology. It combines a theoretical energy function, like a molecular mechanics force field, with terms that represent the experimental observations, such as X-ray diffraction patterns or NMR restraints. experimental refinement model A model of the disordered solvent in a crystal, used during experimental structure refinement to account for the scattering that ordered atoms do not explain. It treats that solvent as a uniform continuum of constant electron density rather than as individual water molecules. Without it, the low-resolution part of a diffraction pattern fits poorly. (UNVERIFIED) bulk solvent model A non-bonded potential energy function modified so that the energy stays finite when 2 atoms overlap completely. It removes the singularity in the Lennard-Jones and Coulomb terms that appears as an atom is created or annihilated in an alchemical calculation. Without it the calculation becomes numerically unstable near the end points of the coupling parameter. (UNVERIFIED) soft core potential A force field in which the total energy is the sum of independent terms, each depending on a fixed set of atoms and unaffected by the rest of the system. Independence is what makes the energy cheap to compute and easy to parallelise. It also means the parameters cannot respond to their environment, which is what a polarizable force field adds. (UNVERIFIED) additive force field A model that describes a system through the probabilities of its possible microscopic arrangements, rather than by following any one of them. It connects what individual particles do to the temperature, pressure and free energy that are actually measured. Every thermodynamic quantity taken from a simulation rests on one. (UNVERIFIED) statistical model Not to be confused with STATO:0000107 "statistical model", which is a statistics and inference concept: a formalization of relationships between random variables. This term is the statistical-MECHANICS sense - the theoretical framework connecting the microscopic motion of atoms to macroscopic thermodynamic properties, for example the canonical ensemble. Renamed 2026-07-29 to remove that ambiguity; "statistical model" is retained as an exact synonym. statistical mechanics model A statistical mechanics model giving how the speeds of particles are spread at a given temperature, when the system is at thermal equilibrium. Most particles move near an average speed, with a long tail of fast ones and few very slow ones. Starting velocities in a simulation are drawn from it, and the observed distribution is checked against it to confirm the system has equilibrated. (UNVERIFIED) Maxwell-Boltzmann distribution A two-state transition model is a simplified statistical mechanics model used to describe a system that can exist in two distinct states, such as a protein in a folded or unfolded state. The model characterizes the probability of switching from one state to the other, often used to analyze the kinetics of conformational changes. two-state transition model An integration grid for density functional theory calculations built by giving each atom its own spherical grid and blending these with weights that sum to 1 everywhere. The blending avoids double counting where atomic grids overlap. It is the standard way such molecular grids are built. (UNVERIFIED) Becke integration grid A set of points and weights on a sphere chosen so that a numerical integration over that sphere is exact for spherical harmonics up to a stated order. It reaches a given accuracy with far fewer points than an evenly spaced grid. It is the angular part of the grid a density functional theory calculation integrates on. (UNVERIFIED) Lebedev integration grid A rigid 3-site water model retuned from SPC/E so that proteins simulated in it rotate at the rates measured experimentally, giving better agreement with NMR relaxation data. Correct rotation matters because it sets the NMR relaxation properties computed from a simulation. It costs the same as SPC/E. (UNVERIFIED) SPC/E-b A rigid 3-site water model refitted from its predecessor against a wide set of measured properties simultaneously, rather than a few. Fitting to many targets at once gives better agreement across temperature and pressure than fitting to a handful. It keeps the 3-site form, so it costs no more than the model it replaces. (UNVERIFIED) TIP3P-FB A rigid 3-site water model whose charges and geometry were optimised together, by a nearly exhaustive search over the electrostatic properties most relevant to liquid water, so that it reproduces a wide set of measured bulk properties. It is significantly more accurate than TIP3P and SPC/E over a wide range of temperatures, at the same cost. It also reproduces the charge hydration asymmetry of water, the different solvation of positive and negative ions of the same size. (UNVERIFIED) OPC3 A polarizable 3-site water model of the OPC family, based on the Drude oscillator idea, whose charge distribution responds to the electric field around it. Real water is markedly more polar in the liquid than as an isolated molecule, and a fixed-charge model must compromise between the 2. It removes that compromise with little extra computational cost, and its parameters were optimised to match 5 key properties of bulk water. (UNVERIFIED) OPC3-pol A coarse-grained water model in which 4 charged beads, joined in a tetrahedron, together represent a cluster of about 11 water molecules. Because its beads carry charges, it produces its own dielectric response instead of needing a uniform dielectric medium. It was developed as the explicit solvent for the SIRAH force field. (UNVERIFIED) WatFour (WT4) coarse-grained A crystallographic model component is a fundamental concept used to describe the structure of a crystal. It represents one of the key pieces of information, such as the repeating unit or the symmetry rules, that define how molecules are arranged in a crystalline solid. crystallographic model component The smallest portion of a crystal from which the whole structure can be built by applying the symmetry operations of its space group. It may hold 1 molecule, several, or a fraction of 1, depending on where the molecules sit relative to the symmetry elements. It is what a deposited crystal structure file usually contains. (UNVERIFIED) asymmetric unit A solvation shell refers to the layer of solvent molecules that are in direct contact with or in the immediate vicinity of a solute molecule. These solvent molecules are often more ordered and have different dynamic properties compared to the bulk solvent further away. Usually, analyzing the structure and dynamics of the first and second solvation shells is crucial for understanding the details of solute-solvent interactions. solvation shell A way of describing the 3-dimensional arrangement of atoms in a molecule, either as Cartesian coordinates or as internal coordinates such as bond lengths, angles and dihedral angles. Both describe the same shape, but internal coordinates leave out where the molecule sits and how it is turned in space, and they make a constraint or a symmetry easy to impose. The choice is usually dictated by the program. (UNVERIFIED) molecular structure representation A way of describing a molecule's geometry by giving each atom's position relative to atoms already placed, as a bond length, an angle and a rotation. It expresses shape directly, so a symmetry or a constraint can be imposed by fixing one number. It becomes unstable when 3 atoms it refers to lie nearly in a straight line. (UNVERIFIED) Z-matrix coordinate representation The collection of fitted numbers that gives a force field its behaviour for a particular class of molecules, such as proteins, lipids or carbohydrates. The functional form stays the same across sets; the numbers change. A simulation states both the form and the set, because the form alone does not identify what was used. (UNVERIFIED) force field parameter set A parameter set covering phosphorylated amino acids such as phosphoserine, phosphothreonine and phosphotyrosine, for use with the AMBER ff99SB protein force field and older AMBER protein force fields. Standard protein parameters do not include them, because phosphorylation is a modification rather than a standard residue. The newer phosaa14SB and phosaa19SB sets replace it for the ff14SB and ff19SB force fields. (UNVERIFIED) phosaa10 parameter set A parameter set covering phosphorylated amino acids such as phosphoserine, phosphothreonine and phosphotyrosine, for use alongside the ff14SB protein force field. Standard protein parameters do not include them, because phosphorylation is a modification rather than a standard residue. It is matched to that protein force field and should not be used with another. (UNVERIFIED) phosaa14SB parameter set A parameter set covering phosphorylated amino acids such as phosphoserine, phosphothreonine and phosphotyrosine, for use alongside the ff19SB protein force field. Standard protein parameters do not include them, because phosphorylation is a modification rather than a standard residue. It is matched to that protein force field and should not be used with another. (UNVERIFIED) phosaa19SB parameter set A force field parameter set for the monovalent alkali and halide ions, published by Joung and Cheatham in 2008, with parameters fitted separately for each of 3 water models. Fitting per water model is the point: parameters that behave correctly with 1 water model do not with another. They were fitted to solvation free energies, ion-water interaction energies and the lattice energies and lattice constants of salt crystals. (UNVERIFIED) Joung/Cheatham ion parameter set A collection of mathematical functions combined to represent the shape of each electron orbital in a quantum chemistry calculation. The true orbitals cannot be written down, so they are built up from a finite set of simpler functions instead. The choice of these functions, together with the quantum method, largely sets both the accuracy and the cost of a calculation. (UNVERIFIED) basis set A set of mathematical functions used to approximate products of orbital basis functions, rather than the orbitals themselves, so that electron repulsion integrals can be computed more cheaply. It is what makes density fitting and resolution-of-identity methods fast. It must be matched to the orbital basis set it is paired with, or the speed-up costs accuracy. (UNVERIFIED) auxiliary basis set A force field parameter set that extends the AMBER Lipid17 and Lipid21 lipid force fields with further head groups, such as lysophospholipids, phosphatidylinositols and cardiolipins, and with sterols such as ergosterol. It reuses the bonded and Lennard-Jones parameters of Lipid17 or Lipid21, with new partial charges fitted by the RESP method over several conformations. Its head groups are still in development and are less well tested than those of the main lipid force field. (UNVERIFIED) Lipid_ext parameter set A QM/MM model component that adds a deep neural network correction to the energy of a QM/MM potential, trained to improve both the interactions within the quantum region and those between quantum atoms and nearby classical atoms. The correction is non-electrostatic and acts only on classical atoms within a cutoff distance of the quantum region, which is what range-corrected refers to. It is typically used to bring a fast semi-empirical QM/MM model close to the accuracy of ab initio QM/MM. (UNVERIFIED) deep potential range corrected potential A description of the full symmetry of a crystal, listing the rotations, reflections, translations and their combinations that leave the structure unchanged. There are 230 such groups in 3 dimensions, and every crystal belongs to exactly 1. Together with the asymmetric unit it determines the entire crystal. (UNVERIFIED) space group A generalized Born implicit solvent model that computes effective Born radii with the scheme of Onufriev, Bashford and Case, which rescales the radii to account for the spaces between atomic spheres that earlier pairwise models missed. Its radii come closer to the true molecular volume at little extra cost. In AMBER it has 2 parameter variants, igb=2 and igb=5, which are recommended for most protein work and are used with the mbondi2 radii. (UNVERIFIED) OBC implicit solvent model A generalized Born implicit solvent model that corrects the effective Born radii for the narrow regions between overlapping atomic spheres, which earlier formulations treated as more solvent-exposed than they are. The correction matters most in crevices and buried pockets. The letter n in the name stands for those neck regions. (UNVERIFIED) GBn implicit solvent model A label classifying an atom by its element and its local bonding environment, in the scheme used by the SYBYL software and by the Mol2 file format. For instance, C.ar marks an aromatic carbon and C.2 a carbon with a double bond. Many modelling tools read these types from Mol2 files, for example when preparing a small molecule for docking or charge calculation. (UNVERIFIED) SYBYL atom type A generalized distance coordinate restraint is a type of restraint that is applied to a complex coordinate defined as a combination of multiple distances, angles, or dihedrals. This allows for the restraining of complex conformational changes, such as the movement of a whole domain, that cannot be described by a single distance. generalized distance coordinate restraint An electronic property giving how a molecule's electrons respond to a magnetic field, as a 3-by-3 array because the response depends on the field's direction relative to the molecule. A molecule with circulating electrons, such as an aromatic ring, responds much more strongly along one axis than the others. It is what underlies the ring current shifts seen in NMR spectra. (UNVERIFIED) magnetic susceptibility tensor An alignment tensor is a mathematical object that describes the degree and direction of the average orientation of molecules in a sample that is not perfectly random. In an NMR experiment, molecules can be weakly aligned by a magnetic field, and the alignment tensor quantifies this small but measurable preference in orientation. This tensor is crucial for interpreting experimental data like residual dipolar couplings (RDCs), which provide powerful long-range structural information. alignment tensor A force field component that is 1 of the numbers setting the Lennard-Jones van der Waals interaction of an atom type, such as its well depth (epsilon), its size (sigma), or the coefficient of the attractive or the repulsive term. Parameters for 2 different atom types are combined by a combining rule. Those combining rules differ between force fields, which is one reason their parameters cannot be mixed. (UNVERIFIED) Lennard-Jones parameters Lennard-Jones parameter An implicit solvent model that represents a solute as a low-dielectric cavity in a high-dielectric continuum and solves for the electrostatic potential, including the screening by mobile salt ions. It makes no analytical approximation to the molecule's shape, so it is the reference the faster models are judged against. Solving it numerically is expensive, which is why it is usually applied to snapshots rather than during a simulation. (UNVERIFIED) Poisson-Boltzmann model A Poisson-Boltzmann model in which the equation is linearized and solved in closed form under simplifying assumptions about geometry. It is far faster than solving the full non-linear equation numerically. It holds only where electrostatic potentials are small, so it is unreliable for highly charged systems. (UNVERIFIED) analytical linearized Poisson-Boltzmann model A generalized Born implicit solvent model that accounts for charge hydration asymmetry, the fact that a negative ion and a positive ion of the same size are not solvated equally by water. It adds this asymmetry to the generalized Born equation to copy the response of a chosen explicit water model, and uses its own set of atomic radii. It was tested on small molecules, amino acid analogues and small proteins, and is available in the GBNSR6 program. (UNVERIFIED) CHAGB implicit solvent model A generalized Born implicit solvent model in which ions are kept as explicit particles while the water is treated implicitly, with extra coefficients in the generalized Born equation that tune the interactions between solute atoms and ions and between ions. Interactions between solute atoms are computed with the standard generalized Born model, and it has been used together with the GB-Neck2 model. It was fitted to explicit-solvent ion distributions around DNA, and captures local effects of ions that a uniform salt concentration cannot. (UNVERIFIED) GBION implicit solvent model A generalized Born implicit solvent model that computes each atom's effective Born radius numerically, by an integral over the molecular surface of the solute (the R6 integral), rather than by an analytical approximation. This makes it nearly as accurate as Born radii derived from Poisson-Boltzmann calculations, and almost independent of the choice of atomic radii. In AMBER it is used to estimate solvation free energies of single structures and cannot yet be used in molecular dynamics. (UNVERIFIED) GBNSR6 implicit solvent model A model that estimates the free energy of a process by adding up tabulated free energies for moving each chemical group between environments, such as from water into a protein interior. The group values are measured once on small molecules and then reused. It is fast and additive, which also means it cannot capture effects that depend on how groups are arranged relative to one another. (UNVERIFIED) group transfer free energy model A database holding the 2-dimensional and 3-dimensional structures of small organic compounds. Such collections supply the starting geometries and the reference conformations that force field parameters are fitted and tested against. They differ from protein structure archives in holding many small entries rather than few large ones. (UNVERIFIED) small molecule structural database A RISM model that solves the integral equations on a 3-dimensional grid around a solute, giving the spatial distribution of each solvent site. It shows where solvent molecules are likely to sit, including individual ordered water positions, without running a simulation. It needs the bulk solvent correlation functions from a 1-dimensional calculation as input. (UNVERIFIED) 3D-RISM model An approximate equation relating the 2 unknown correlation functions in an integral equation theory of liquids, so that the system can be solved. Without it there are 2 unknowns and 1 equation. Different choices, such as hypernetted-chain or Kovalenko-Hirata, give measurably different solvation free energies from otherwise identical input. (UNVERIFIED) closure relation The Kovalenko-Hirata (KH) closure is a mathematical approximation used in the statistical mechanical Ornstein-Zernike (OZ) integral equation theory (commonly utilized in 3D-RISM theory) to calculate solvent structure and solvation free energy. Kovalenko-Hirata closure An implicit solvent model in which the solvent excluded surface, the boundary between solute and solvent used in a Poisson-Boltzmann calculation, is predicted by a machine-learned model instead of being built geometrically. It produces the surface much faster than the classical construction while closely reproducing it. In AMBER it is used within Poisson-Boltzmann calculations through models such as GENIUSES. (UNVERIFIED) machine-learned solvent excluded surface model A machine-learned potential that predicts only the difference between a cheap approximate calculation and an accurate one, rather than the full energy. The difference is smaller and smoother than the energy itself, so far less training data is needed to learn it well. The cheap calculation still has to be run, which is the cost of the approach. (UNVERIFIED) delta machine learning potential A charge model that fits permanent point charges to the electrostatic potential and adds an induced dipole on each atom to represent the response to a field. The permanent part describes the isolated molecule and the induced part its response to the environment. It is 1 way of deriving the electrostatic parameters of a polarizable force field from quantum calculations. (UNVERIFIED) RESP induced dipole charge model A charge model that fits a dipole and higher terms on each atom, not only a point charge, to reproduce the electrostatic potential around a molecule. The extra terms capture features a point charge cannot, such as the anisotropy of a lone pair or the positive region on a halogen. It is more accurate and more expensive to use. (UNVERIFIED) RESP permanent multipole charge model The AMBER FB15 (force-balanced 15) protein force field is a refinement of the AMBER94 ff with a systematic and thorough optimization of the intramolecular terms. The key difference in the optimized result is a significant lowering of the potential in regions away from the energy minima, which is expected to yield greater flexibility in finite temperature simulations. This force field is particularly useful for simulations of proteins, particularly when fluctuations away from equilibrium, conformational changes, and/or temperature dependence are expected to play important roles. AMBER fb15 A force field in which non-polar hydrogens are merged into the heavy atom they are bonded to, so that a methyl group is 1 interaction site. This reduces the number of particles, so a simulation costs less than with an all-atom model. Polar hydrogens are kept as separate sites, because they take part in hydrogen bonds. (UNVERIFIED) united-atom force field A charged particle attached by a harmonic spring to an atom in a polarizable force field, used to represent how that atom's electron cloud shifts in an electric field. The local electric field pulls the particle and its atom apart, and the 2 charges together form an induced dipole, so polarization is reproduced without modelling electrons. The particle is either given a small mass and moved with the other atoms, usually with its own thermostat held at a very low temperature, or its position is solved for at every step. (UNVERIFIED) drude oscillator A RISM model that solves the integral equations for a homogeneous fluid, giving site-to-site correlation functions that depend only on distance. It describes the structure of the bulk solvent itself, with no solute present. Its output is the required input for a 3-dimensional calculation around a solute. (UNVERIFIED) 1D-RISM model A statistical mechanical model of molecular liquids that predicts how solvent molecules arrange themselves around a solute by solving integral equations for site-to-site correlation functions. It gives an atomistic picture of solvation without simulating individual solvent molecules moving. It sits between an implicit solvent model and explicit solvent in both cost and detail. (UNVERIFIED) reference interaction site model RISM model A RISM model corrected so that its results reproduce the known dielectric constant of the solvent. Standard RISM does not, which distorts long-range electrostatics. The correction matters most for solvation free energies of charged solutes. (UNVERIFIED) dielectrically consistent RISM model A thermodynamic property giving the average kinetic energy of the particles in a system, expressed on a conventional scale. In a simulation it is computed from the velocities and fluctuates strongly, because the system is small. A value quoted from a short stretch of trajectory can differ from the target by several degrees for that reason alone. (UNVERIFIED) temperature A QM/MM model component that replaces the fixed Lennard-Jones terms between quantum and classical atoms with exchange repulsion and dispersion terms that depend on the current partial charge of each quantum atom. An atom that loses electron density becomes smaller and less polarizable, and one that gains it becomes larger, which fixed Lennard-Jones parameters cannot represent. It matters most in reactions, where atoms change their charge and bonding as the reaction proceeds. (UNVERIFIED) charge-dependent van der Waals correction A semi-empirical extended tight binding method, the first of its family, parameterised to give good molecular geometries, vibrational frequencies and non-covalent interaction energies for elements across most of the periodic table. Its name records those 3 targets: geometries, frequencies and non-covalent interactions. It is used to optimise structures and screen conformations quickly. (UNVERIFIED) GFN1-xTB A semi-empirical tight binding method that adds atom-centred multipoles and a density-dependent dispersion treatment to the earlier method in its family. Those additions improve the description of non-covalent interactions without the separate hydrogen bond and halogen bond corrections that the earlier method needed. It is the member of the family in general use. (UNVERIFIED) GFN2-xTB A correction to the PM3 semi-empirical quantum method that replaces the core-core repulsion between atoms of different molecules with specially fitted functions, to improve the interactions of organic molecules with water and of water molecules with each other. Separate parameters are used for water-water pairs and for organic-organic and organic-water pairs, and a later version treats hydrogens that attract water differently from hydrogens that repel it. In AMBER it is available in semi-empirical Born-Oppenheimer molecular dynamics. (UNVERIFIED) PIF correction A correction to the PM3 semi-empirical quantum method in which the core-core repulsion functions, within and between molecules, are replaced by specially fitted functions, to improve the description of hydrogen-bonded systems such as liquid water. Its first version has parameters only for hydrogen and oxygen, and a second version adds chlorine. In AMBER it is available in semi-empirical Born-Oppenheimer molecular dynamics. (UNVERIFIED) MAIS correction A force field that allows chemical bonds to form and break during a simulation, so the list of bonded atoms is not fixed in advance. Bond-order force fields such as ReaxFF do this by computing the strength of each bond from the current distances between atoms, while other approaches, such as the empirical valence bond method, mix the energies of several fixed bonding patterns. It can describe reactions in systems far larger than a quantum method can handle, at a cost well above that of a conventional force field. (UNVERIFIED) reactive force field A reactive force field in which the strength of every bond is recomputed from interatomic distances at each step, so bonds can form and break. Charges are also recomputed, since they change as bonds do. It reaches system sizes no quantum method can while still describing reactions, at a cost much higher than that of a conventional force field. (UNVERIFIED) ReaxFF A force field component that replaces the Lennard-Jones parameters for 1 particular pair of atom types, overriding the values the combining rule would give, as done in the CHARMM force fields. It is used where the combining rule gives poor results for a pair, for example for some ions with carboxylate oxygens. It is applied pair by pair rather than by changing the atom types themselves. (UNVERIFIED) NBFIX correction An algorithm that computes the induced dipole on each atom in a polarizable force field. Each induced dipole changes the field felt by the others, so they must be solved together rather than one at a time. That coupling is what makes a polarizable simulation several times more expensive than a fixed-charge one. (UNVERIFIED) polarization algorithm A polarization algorithm that computes the induced dipoles in a set of polarizable atoms by giving each atom a point polarizability and solving the resulting linear equations self-consistently. It captures how the dipole induced on one atom polarizes its neighbours in turn. Its induced dipoles can grow without limit when 2 atoms come too close, a problem called the polarization catastrophe, which later schemes such as Thole's avoid by damping the interaction at short range. (UNVERIFIED) Applequist scheme A charge model that fits point charges to the electrostatic potential while allowing some of them to sit at fixed positions away from any nucleus. Placing charge off-nucleus reproduces directional features, such as a lone pair, without the cost of higher multipole terms. It is the cheaper alternative to a multipole description. (UNVERIFIED) RESP permanent virtual site charge model A potential energy function formed as a straight-line interpolation between the functions of an initial and a final state, controlled by a coupling parameter running from 0 to 1. It is what makes a continuous path between 2 chemically different states possible in an alchemical calculation. It behaves badly when atoms appear or disappear, which is why the soft-core variant exists. (UNVERIFIED) linear mixing potential A data format defining, for each residue that can gain or lose a proton, the topology and charges of every protonation state it can take and the reference energy of each. A constant pH simulation reads it to know what states are available and what switching between them costs. A residue absent from the library cannot titrate. (UNVERIFIED) CpHMD library format A force field component that adds an energy correction read from a 2-dimensional grid indexed by a pair of neighbouring backbone rotation angles. The pair of angles is treated together rather than as 2 independent terms, which is what lets the correction fix their coupling. It was introduced to bring simulated backbone conformational populations into line with experiment. (UNVERIFIED) CMAP correction map A restraint potential is an artificial energy function that is added to the force field to guide a simulation. This potential adds an energy penalty when the system deviates from a user-defined geometric target, such as a specific distance, angle, or RMSD value. restraint potential A statistical mechanical theory that describes solvent around a solute by treating the solvent as a fluid whose density and orientation vary with position. It assigns energy and entropy to each region rather than to the system as a whole. That local assignment is what lets individual water sites be scored for displacement by a ligand. (UNVERIFIED) inhomogeneous fluid solvation theory A simulation technique in which the box is treated as one cell of an infinite lattice of identical copies, so a particle leaving one side re-enters the opposite side. It removes the surface that a finite box would otherwise have, where molecules would behave unlike those in bulk. The box must be large enough that a molecule does not interact with its own image. (UNVERIFIED) periodic boundary conditions An expanded ensemble simulation is an enhanced sampling method where the simulation samples not only the coordinates of the system but also an additional thermodynamic variable, such as temperature or an alchemical coupling parameter. The system can move dynamically between different temperatures or states, which helps it to overcome energy barriers and sample a wider range of conformations. For experts, this method is a powerful framework that generalizes techniques like replica-exchange and lambda-dynamics for calculating free energies. expanded ensemble simulation A software program that assembles the parts of a simulation into a system that is ready to run. It places the molecule in a box, fills the box with solvent, adds ions to reach the intended charge and concentration, and writes the topology and coordinate files. It is the step at which a structure becomes a simulatable system. (UNVERIFIED) system setup tool A system preparation program in AmberTools that adds a stated number of copies of a solvent molecule or ion to a crystal unit cell or supercell, placing each at a random position and orientation where it does not clash with the biomolecule. Added molecules are allowed closer to one another than to the biomolecule, because enough solvent must be added to keep the cell at its crystallographic size. It is part of the AmberTools route for building simulations of biomolecular crystals. (UNVERIFIED) AddToBox An electrostatics analysis tool is a type of software designed to calculate and visualize the electrostatic properties of a molecule. These tools typically take a molecular structure with assigned atomic charges and compute quantities like the electrostatic potential on the molecule's surface or the total electrostatic energy of solvation. They are essential for understanding how a molecule will interact with other charged molecules and its polar solvent environment. electrostatics analysis tool A numerical solver that computes the electrostatic potential around a molecule in solvent by solving the Poisson-Boltzmann equation on a grid that is refined where the potential changes fastest. Refining only where needed gives the accuracy of a fine grid at much lower cost. It is the usual way continuum electrostatics is computed for a large biomolecule. (UNVERIFIED) adaptive Poisson-Boltzmann solver A set of van der Waals radii that is the third modification of the Bondi set in AMBER, with further adjustments to the radii of some atoms based on protein simulations. It is the radii set recommended for the GB-Neck2 generalized Born model (igb=8). For nucleic acids its radii are currently the same as mbondi2. (UNVERIFIED) mbondi3 An analysis program that processes the output of a constant pH molecular dynamics run and reports pKa values and titration curves for the titratable residues. It reads the record of protonation state changes rather than the trajectory. It is what turns a constant pH simulation into a number that can be compared with experiment. (UNVERIFIED) cphstats A defined set of functions and rules through which one piece of software uses another. It fixes what can be asked for and what comes back, so the 2 sides can change independently as long as the agreement holds. It is how a simulation engine is driven from a script rather than from a command line. (UNVERIFIED) application programming interface A programming platform and interface for running general computation on NVIDIA graphics processors. A graphics processor performs thousands of simple operations at once, which suits computing forces on many atoms simultaneously. Many simulation engines use it to run several times faster than on central processors alone, at the cost of working only on one manufacturer's hardware. (UNVERIFIED) CUDA A software program that computes free energy differences from the data produced by molecular simulations. It implements estimators such as the weighted histogram analysis method or the multi-state Bennett acceptance ratio, which combine data from many biased or alchemical runs, or endpoint methods that use only the start and end states. It is usually the last step of a free energy workflow, turning the output of many runs into 1 number. (UNVERIFIED) free energy analysis tool An analysis program, part of the FE-ToolKit library in AmberTools, that computes the free energy of 1 alchemical transformation, an edge, in 1 or 2 environments using the multi-state Bennett acceptance ratio and thermodynamic integration. It writes a report with error estimates, time series and convergence checks, and a breakdown of the relative free energy into its parts. The results for many edges are then combined in a network analysis by a companion script. (UNVERIFIED) edgembar A software library is a collection of pre-written code and subroutines that provide specific functionalities to be used by other programs. Libraries are not executed directly by end-users but are imported or linked by developers to perform specialized tasks, such as mathematical computations or file parsing. library software library A software library for computing discrete Fourier transforms, which convert data between a description in space or time and one in frequency. Particle mesh Ewald performs one at every step, so its speed sets a large part of a simulation's cost. The library measures the machine it is running on and chooses its method accordingly. (UNVERIFIED) FFTW A simulation engine distributed with AmberTools that is now used mainly for force field parameter development, including fitting bonded terms and CMAP surfaces and deriving IPolQ charges. It began as a test platform for new molecular dynamics algorithms and keeps a modest ability to run basic simulations, including many small generalized Born simulations at once for validating parameters. It is a development tool rather than a production engine. (UNVERIFIED) mdgx A small script that performs one well-defined supporting task inside a larger workflow, such as converting between file formats or pulling one value out of an output file. It does no science itself; it makes the pieces that do fit together. A published workflow usually depends on several of them. (UNVERIFIED) utility script A utility script in AmberTools, part of the pyMSMT toolbox, that converts a PDB file holding a single metal ion into a mol2 file carrying the ion's charge, which the user specifies. It was written for users of the metal centre parameter builder MCPB.py, which needs the metal ions as mol2 files. (UNVERIFIED) metalpdb2mol2.py A set of atomic radii fitted for use with the generalized Born model of the same authors. Radii and generalized Born model are fitted together, so using 1 with a different model degrades both. It is the default radii set for that model in AMBER. (UNVERIFIED) Onufriev-Bashford-Case radii set The ndfes software, included in the FE-ToolKit package, is used to analyze biased umbrella sampling to estimate free energy surfaces in an arbitrary number of reaction coordinate dimensions using either the variational free energy profile (vFEP), multistate Bennett acceptance ratio method (MBAR), or the highly-related weighted thermodynamic perturbation and generalized weighted thermodynamic perturbation methods. ndfes ParmEd is a software package and Python library designed for manipulating and converting molecular topology and parameter files from various simulation packages, with a primary focus on the AMBER force field format. It allows users to programmatically modify force field parameters, change molecular connectivity, and convert between different file formats. ParmEd A system preparation program that cleans a Protein Data Bank structure file so the AMBER tools can read it. It renumbers residues, corrects atom naming, and reports problems such as missing atoms, alternate locations and unusual residues. It is run before force field parameters are assigned, because those steps fail on an unprepared file. (UNVERIFIED) pdb4amber A system setup program in AmberTools that builds a larger simulation cell by repeating a crystallographic unit cell along chosen directions. It is needed when the unit cell is too thin for the cut-off distance used for non-bonded interactions, because a periodic simulation cell must be able to hold a sphere of at least that radius plus a small buffer. Repeating the cell makes the box big enough, and can also reduce artefacts from the strict symmetry that periodic copies impose. (UNVERIFIED) PropPDB A utility script in AmberTools, part of the pyMSMT toolbox, that cuts a segment out of a protein structure and caps the cut ends, producing clusters suitable for quantum calculations. The user chooses which residues to keep whole, which to keep only as side chains and which to cap as acetyl or N-methyl groups, and it writes Gaussian input files. (UNVERIFIED) ProScrs.py A utility program in AmberTools, written in Python to replace the older resp program, that fits atomic charges to an electrostatic potential computed by quantum chemistry, with restraints that keep poorly determined charges small. Besides the standard RESP charge model for additive force fields, it fits models with induced and permanent dipoles for polarizable force fields. (UNVERIFIED) py_resp.py A utility program in AmberTools that writes the input files for the 2 stages of a restrained electrostatic potential fit with resp or py_resp.py, starting from the electrostatic potential file made by espgen. It records which charges, and for polarizable models which permanent dipoles, must be equal, and the restraint weights for each stage. It performs no fitting itself. (UNVERIFIED) pyresp_gen.py Pytraj is a Python library that provides a programmable interface to the powerful trajectory analysis capabilities of the cpptraj engine from the AmberTools suite. It allows users to write Python scripts to perform a wide range of analyses, such as calculating RMSD, distances, and hydrogen bonds, directly on simulation trajectories. Pytraj enables the creation of complex, automated, and reproducible analysis workflows that integrate seamlessly with other Python data science tools. pytraj A quantum chemistry engine in AmberTools that performs semi-empirical calculations, either on its own or as the quantum part of a QM/MM simulation run with sander. It implements methods such as AM1, PM3, PM6 and MNDO, and the density functional tight binding methods DFTB2 and DFTB3. Antechamber calls it to compute the AM1 charges from which AM1-BCC charges are derived. (UNVERIFIED) SQM A quantum chemistry engine for Hartree-Fock and density functional theory calculations that was written from the outset to run on graphics processors rather than adapted to them later. Designing for that hardware gives it a large speed advantage on the calculations it supports, which suits very large molecules and the quantum part of QM/MM simulations. It supports fewer methods than a general-purpose package. (UNVERIFIED) TeraChem A software library used by the AmberTools PBSA program, built on LibTorch, the C++ form of the PyTorch machine learning framework, to run trained machine learning models for parts of a Poisson-Boltzmann surface area calculation, such as building the molecular surface. It lets those models run on central or graphics processors without a Python interpreter. The options that need it work only when LibTorch is installed. (UNVERIFIED) torch PBSA runtime A system setup program that rebuilds a complete crystallographic repeating unit from a structure file by applying the symmetry operations recorded in it. A deposited file usually contains only the smallest unique part, leaving the rest to be generated. The program performs that generation so the full contents of the cell can be simulated. (UNVERIFIED) UnitCell A graphical system preparation program that builds a molecular system for AMBER interactively: loading a structure, applying a force field, adding solvent and ions, and writing the topology and coordinate files. It is the graphical front end to the same engine the command-line tool uses. The 2 produce identical output from identical commands. (UNVERIFIED) xleap A program that controls how source code is turned into a working executable, working out what must be compiled, in what order and with which options, and calling the compiler and linker to do it. It also finds the libraries the program depends on. Which options it was given affects how fast the resulting program runs, and sometimes what it can do. (UNVERIFIED) build system tool A build system tool that reads a single description of how a program is assembled and generates the native build files, such as Makefiles or Visual Studio projects, for each platform. It lets one set of instructions serve Linux, macOS and Windows. Many simulation packages, including GROMACS and AMBER, are built with it, so which version was used is part of reproducing an installation. (UNVERIFIED) CMake The Intel oneAPI Math Kernel Library (oneMKL) is a collection of highly optimized mathematical functions for science, engineering, and financial applications, specifically tuned for Intel processors. It provides high-performance implementations of routines for linear algebra (BLAS, LAPACK, ScaLAPACK), Fast Fourier Transforms (FFT), and other core mathematical operations. For biomolecular simulation software, linking against oneMKL can significantly boost performance by accelerating the fundamental mathematical computations underlying the simulation algorithms. Intel oneAPI Math Kernel Library A package manager is a tool that automates the process of installing, upgrading, configuring, and removing software packages and their dependencies. It simplifies the management of the complex web of software required for scientific computing by handling the details of where to find packages and which versions are compatible. For simulation scientists, package managers like Conda are crucial for creating stable and reproducible software environments. package manager Miniconda is a minimal, bootstrap installer for the Conda package manager, which is widely used in the scientific computing community. It includes only the Conda package manager and Python, allowing users to create customized software environments by installing only the packages they need. For researchers, Miniconda provides a lightweight way to set up isolated and reproducible environments for their simulation and analysis software. Miniconda A utility script distributed with AmberTools that fetches and applies official updates and fixes to an existing installation. It contacts the update server, downloads what applies to the installed version, and patches the source. Which updates have been applied is part of what identifies the version a result came from. (UNVERIFIED) update_amber An analysis program that solves the 1-dimensional reference interaction site model (RISM) equations for a pure solvent, giving its structural and thermodynamic properties. Its output describes the bulk liquid with no solute present. It must be run before any 3-dimensional RISM calculation, which takes those results as input. (UNVERIFIED) rism1d A quantum chemistry engine for Hartree-Fock and density functional theory calculations, distributed with AmberTools, that runs in parallel on central processors and on NVIDIA and AMD graphics processors. Evaluating the electron repulsion integrals, where most of the computing time goes, on graphics processors gives it its speed. It serves as the quantum part of QM/MM simulations with AMBER, where the calculation must be repeated at every step. (UNVERIFIED) QUICK A quantum chemistry program for very high accuracy electronic structure calculations, implementing coupled cluster methods to arbitrary excitation level and their multi-reference forms. Those methods are the benchmark other methods are measured against. Its most exact methods are limited to small molecules, but its local natural orbital approximations bring coupled cluster accuracy to molecules of hundreds of atoms. (UNVERIFIED) MRCC A quantum chemistry engine that performs density functional calculations using orbitals confined to a fixed radius around each atom, with the resulting integrals tabulated in advance. Confining the orbitals is what makes the tabulation possible and the calculation fast. It is used for large systems where full density functional theory would be too slow. (UNVERIFIED) Fireball A quantum chemistry engine implementing density functional tight binding, an approximation that replaces most of the electronic integrals with values tabulated in advance. It is 2 to 3 orders of magnitude faster than full density functional theory while remaining quantum mechanical. Its accuracy depends on parameters fitted for each pair of elements, so it works only for element combinations that have been parameterised. (UNVERIFIED) DFTB+ A quantum chemistry program implementing the GFN family of extended tight binding methods, semi-empirical methods parameterised for most of the periodic table. It is fast enough to optimise geometries, compute vibrational frequencies and search conformations for molecules of hundreds of atoms. It is widely used as a preliminary step before a more expensive calculation. (UNVERIFIED) xTB A software framework that connects separate simulation programs so that each can compute the part of a system it is best suited to. It passes coordinates, energies and forces between them and keeps them in step. It is what allows a quantum chemistry program and a molecular dynamics program to run 1 simulation together. (UNVERIFIED) PUPIL A software library that lets a classical simulation program request energies and forces from the TeraChem quantum chemistry program over a network connection. The 2 programs run separately and exchange data at each step. It is what allows a quantum and a classical engine to run one hybrid simulation together. (UNVERIFIED) TCPB-cpp A simulation engine built to run reactive force field simulations, where bonding is recomputed at every step rather than fixed. Recomputing the bond network is what makes such a simulation expensive, and the engine is written around doing it in parallel. It is used for reactions in materials and at surfaces. (UNVERIFIED) PuReMD A structure preparation program that adds hydrogen atoms to a macromolecular structure and chooses their positions. It also decides how to orient groups whose 2 possible arrangements X-ray data cannot distinguish, such as the ends of asparagine and glutamine side chains. Crystallography rarely sees hydrogens, so a structure file usually arrives without them and a simulation cannot start until they are there. (UNVERIFIED) reduce A system setup utility that places the hydrogen atoms of water molecules taken from an experimental structure. Crystallography sees the oxygen of an ordered water and not its hydrogens, so their orientation must be worked out. It chooses orientations that make sensible hydrogen bonds with the surroundings. (UNVERIFIED) guess water hydrogens gwh A Python library and set of command-line tools for analysing and manipulating Protein Data Bank structure files and for running molecular dynamics simulations with AMBER, GROMACS and NAMD through 1 interface. It covers structure preparation, minimization, dynamics and trajectory processing, and PACKMOL-Memgen uses it to estimate the volume of a protein. (UNVERIFIED) pdbremix A system preparation program that builds an initial arrangement of molecules inside a defined region by packing them to satisfy stated counts and spatial constraints while keeping them from overlapping. It solves a constrained optimisation rather than placing molecules at random. It is how a mixed or layered starting system is built when a regular lattice will not do. (UNVERIFIED) Packmol AMBAT is a software tool used to construct three-dimensional models of cell membranes for use in AMBER molecular simulations. Consisting of a set of scripts, it automates the complex process of arranging lipid molecules to form a proper bilayer and can then embed a protein structure within that prepared membrane. As part of the AmberTools suite, AMBAT (Amber Membrane Builder and Analysis Tool) provides a script-based workflow for both the initial construction of lipid bilayer systems and for subsequent analysis of membrane properties, serving as a specialized tool for preparing membrane protein simulations. Amber Membrane Builder and Analysis Tool AMBAT A software program that derives force field parameters by adjusting them until the force field reproduces reference data, usually quantum calculations. It is used when no existing parameter covers a molecule and assignment by analogy is not good enough. Its output is a parameter set, not a prepared system. (UNVERIFIED) parameter fitting tool A utility program that searches automatically for force field parameters giving the best agreement with reference data, by trying many combinations rather than fitting one term at a time. Searching the whole space at once can find a combination that fitting terms separately misses. It is used for parameter development rather than for preparing a single system. (UNVERIFIED) described in https://doi.org/10.1002/jcc.1079 Parmscan A utility program in AmberTools that checks whether every parameter a molecule needs is present in the chosen force field, and writes an additional parameter file supplying the missing ones by analogy. Missing parameters otherwise stop system preparation with an error. Parameters it supplies by analogy are guesses and are flagged as such in its output. (UNVERIFIED) parmchk2 An interactive utility included in the AmberTools software package for deriving force field parameters. It assists users in calculating new parameters for bond lengths and angles based on a set of built-in empirical rules. This tool is particularly useful when parameterizing a new molecule for which no existing force field parameters are available. (UNVERIFIED) parmcal A utility program in AmberTools that optimizes chosen force field parameters, often dihedral terms, so that the energies or forces the force field gives over many conformations of a molecule match quantum chemical reference values. It is used when parameters are missing and cannot be supplied by analogy, or when existing ones are not accurate enough. It can also fit 1 set of parameters across several molecules at once. (UNVERIFIED) paramfit A utility program in AmberTools that builds force field parameters for a metal centre in a protein, from quantum calculations on the metal and the groups around it. A metal site has no standard parameters, because its geometry and charge depend on which groups coordinate it. It automates a procedure that is otherwise long and easy to get wrong. (UNVERIFIED) MCPB.py A Python toolbox in AmberTools for modelling metal sites in biomolecules for molecular dynamics, covering both bonded and non-bonded models of the metal. It bundles several programs, including the metal centre parameter builder MCPB.py, the ion parameterization program IPMach.py and the metal site search program PdbSearcher.py. It is a toolbox rather than a single program. (UNVERIFIED) pyMSMT A utility program in AmberTools that assigns a force field atom type to every atom of a molecule, from its element and its bonding environment. Every parameter a molecule receives follows from those assignments. It is run before parameters are looked up, and its output is worth inspecting for unusual chemistry. (UNVERIFIED) atomtype A utility program in AmberTools that turns AM1 Mulliken charges, computed beforehand by a fast semi-empirical calculation, into AM1-BCC charges by adding tabulated bond charge corrections that depend on atom and bond types. The corrections are fitted so the result approximates charges fitted to a far more expensive quantum calculation of the electrostatic potential. Antechamber calls it automatically when these charges are requested. (UNVERIFIED) am1bcc A utility program in AmberTools that assigns each bond in a molecule 1 of the 6 bond types defined for AM1-BCC charges: single, double, triple, aromatic single, aromatic double and delocalized. It works from the connectivity table alone, or adjusts simple bond types already present in the input file. Antechamber uses the result when it applies the AM1-BCC bond charge corrections. (UNVERIFIED) bondtype A utility program in AmberTools that writes the input files for the 2 stages of a restrained electrostatic potential (RESP) fit of atomic charges. It detects which atoms are equivalent and must receive equal charges, and records the restraint weights and any charges or charge groups the user wants fixed. The fit itself is performed by a separate program. (UNVERIFIED) respgen A utility program in AmberTools that writes a residue template, a prep file, for a new molecule from a file holding its structure, atom types and charges. It builds the main chain and the internal coordinate tree itself, or takes the main chain atoms, the head and tail atoms that join neighbouring residues, and atoms to delete from a main chain file given by the user. It is run once per new building block, and its output is reused thereafter. (UNVERIFIED) prepgen A utility program in AmberTools that builds the residue template, in prepi format, for a modified amino acid, nucleotide or other polymer unit, removing the capping groups used during the charge calculation. It runs a sequence of antechamber programs from a single input file, including a RESP charge fit over 1 or more conformations. It is meant for cases too complex for prepgen alone, such as charges averaged over several conformations. (UNVERIFIED) residuegen A quantum chemistry engine that divides a large molecule into overlapping regions, solves each separately, and assembles the result. Doing so makes the cost grow in proportion to the system size rather than far faster, which is what puts whole proteins within reach. The overlap between regions is what keeps the joins from introducing errors. (UNVERIFIED) divcon A diagnostic utility program in AmberTools that checks a small molecule structure file for problems that would make the antechamber programs fail, such as unusual elements, unfilled valences, atom clashes, doubtful bonds and atoms not linked to the rest of the molecule. It then tries to assign bond and atom types and reports any errors and warnings. Since AmberTools 17 the same checks run inside antechamber by default. (UNVERIFIED) acdoctor A utility script in AmberTools, part of the pyMSMT toolbox, that fits Lennard-Jones parameters for an ion, in the 12-6 or 12-6-4 model, to a target hydration free energy, a target ion-oxygen distance, or both. It runs thermodynamic integration simulations of the ion in a chosen water model at trial parameters and adjusts them until the targets are met within a tolerance. (UNVERIFIED) IPMach.py A utility script in AmberTools, part of the pyMSMT toolbox, that searches a list of PDB files for the metal centres of a chosen element. It records each centre's ligating atoms, bond distances and geometry in an environment file and a summary file. It is the Python version of the earlier Pdbsearcher program and recognises metal ions more reliably from PDB naming. (UNVERIFIED) PdbSearcher.py A system preparation script in AmberTools that writes the input files a replica exchange molecular dynamics run needs: 1 input file per replica, the group file listing them and, for a multidimensional run, the file describing each exchange dimension. It builds them from a reference input file and lists of the temperatures, pH values, redox potentials or Hamiltonians to be exchanged. A replica exchange run needs as many input files as it has replicas, which is too many to write reliably by hand. (UNVERIFIED) genremdinputs.py The nfe-umbrella-slice script is a utility used in the analysis of umbrella sampling simulations, likely as part of a larger free energy calculation workflow. Its purpose is to extract a specific portion or "slice" of the simulation data from the collective output of all umbrella windows. Experts use this for tasks such as assessing the convergence of the free energy profile by analyzing how it changes as more data is included. nfe-umbrella-slice A system preparation script, part of AmberTools, that writes the input file a constant pH molecular dynamics run needs. It records which residues may change protonation state and the parameters the titration algorithm uses. Without it the simulation has no list of titratable sites. (UNVERIFIED) cpinutil.py The ceinutil.py script is a utility used to prepare the input files for constant redox-potential MD simulations in AMBER. It allows the user to specify which residues can change their redox state and defines the parameters for the simulation. This script is essential for setting up simulations that model the behavior of molecules in a specific electrochemical environment. ceinutil.py A system preparation script that writes the input file for a simulation holding both pH and redox potential constant. It records which residues may change protonation state and which may change oxidation state. It is the constant pH preparation script extended to redox-active sites. (UNVERIFIED) cpeinutil.py The cyana_rest_to_amberRST script iscript converts CYANA .upl upper distance limit files into AMBER .RST format, allowing for the inclusion of restraints in molecular dynamics simulations. This conversion tool is commonly used to prepare input data for AMBER's sander or pmemd applications. cyana_rest_to_amberRST This script is a utility for creating a restraint file (RST) for a simulation that uses direct dipolar coupling (DIP) data from NMR experiments. It likely reads experimental dipolar coupling values and generates the corresponding energy restraint terms in a format that can be read by a simulation program like AMBER. This automates the process of incorporating this specific type of experimental data into a simulation. makeDIP_RST A utility script in AmberTools, part of FRETrest, that adds pseudo atoms to a structure or topology file to stand for the mean positions of fluorescent dyes, as found by accessible volume simulations. It is the first of 2 steps in adding Förster resonance energy transfer (FRET) restraints to a simulation; FRETrest.py then tethers the pseudo atoms and applies the restraints. It reads a FRET configuration file that names the labelling positions. (UNVERIFIED) placeAV.py The FRETrest.py script is a utility used to generate distance restraints for a molecular dynamics simulation based on experimental data from Förster Resonance Energy Transfer (FRET) experiments. It translates the efficiency of energy transfer between two fluorescent labels into a distance distribution that can be used to bias the simulation. FRETrest.py The ambpdb utility is a command-line tool in the AmberTools suite that converts AMBER's native topology (.prmtop) and coordinate (.inpcrd or .rst7) files into the standard Protein Data Bank (PDB) format. ambpdb A system preparation script that derives the separate receptor and ligand topology files needed for an end-point binding free energy calculation by stripping atoms from the topology of the complex. Deriving them this way guarantees the 3 topologies are consistent. Building them independently is the common source of mismatched-topology errors. (UNVERIFIED) ante-MMPBSA.py A Perl script, part of the AmberTools free energy workflow FEW, that collects the binding free energies and their main energy contributions from MM-PBSA or MM-GBSA calculations on several ligands into 1 table. For each ligand it lists the electrostatic, van der Waals, nonpolar solvation and polar solvation contributions and the total binding free energy. It does not handle the hybrid PB2 model or energy decomposition runs. (UNVERIFIED) extract_WAMMenergies.pl A Perl script, part of the AmberTools free energy workflow FEW, that collects the differences in electrostatic and van der Waals interaction energies from the linear interaction energy results of several ligands into 1 table. It reads each ligand's result file together with a list of ligand names and their measured binding data. The table can then be used to fit the coefficients of a linear interaction energy model. (UNVERIFIED) extract_LIEenergies.pl A Perl script, part of the AmberTools free energy workflow FEW, that chooses which pairs of ligands in a series to connect by thermodynamic integration so that the transformations need, overall, the smallest structural changes. It reads a matrix of pairwise similarity scores, for example from shape overlays with ROCS, and applies Kruskal's algorithm. Relative binding free energies are then computed between the most similar pairs. (UNVERIFIED) identify_transformations.pl A utility script in AmberTools that prepares trajectories from the decharging step of an alchemical binding free energy calculation for BAR/PBSA analysis, which adds the electronic polarization that additive force fields leave out. It strips solvent and ions, writes Poisson-Boltzmann input files with chosen atomic radii and solute dielectric constant, and runs the calculations in parallel with sander. The analysis is used first to calibrate ligand and protein radii and then to scan the dielectric constant of the solute. (UNVERIFIED) bar_pbsa.py A utility script that converts output files from AMBER simulations into the specific data format required by the edgembar free energy analysis program. It processes AMBER's energy or log files to extract the necessary potential energy values for each frame of the simulation. This script is an essential data preparation step for using the edgembar tool to analyze free energy calculations that were performed with the AMBER software. (UNVERIFIED) edgembar-amber2dats.py A utility script in the FE-ToolKit library that performs the free energy analysis of a whole network of alchemical transformations, combining the results edgembar wrote for each edge into a consistent set of relative free energies for all molecules. It can hold chosen edges to experimental values, leave edges out and pick a reference molecule, and writes the analysis as an HTML report. (UNVERIFIED) edgembar-WriteGraphHtml.py A utility script in the FE-ToolKit library that extracts the data edgembar needs from a sander simulation that transforms the system step by step between a molecular mechanics and a QM/MM description. It writes energy time series for multi-state Bennett acceptance ratio and thermodynamic integration analysis, or the nonequilibrium work values when the simulation used Jarzynski sampling. It works like edgembar-amber2dats.py, which handles ordinary alchemical simulations. (UNVERIFIED) edgembar-bookend2dats.py A utility script in the FE-ToolKit library that chooses the number and spacing of the coupling parameter values, the lambda schedule, for an alchemical free energy calculation. It analyses short preliminary simulations run with many states and places the new states so that neighbouring states overlap equally, measured by phase space overlap or predicted replica exchange acceptance. Spacing chosen this way avoids wasting simulations where neighbouring states overlap strongly and leaving gaps where they barely overlap. (UNVERIFIED) fetkutils-tischedule.py A utility script in the FE-ToolKit library that prepares the output of 1 AMBER umbrella sampling simulation for analysis with ndfes. It writes a new file of the restrained coordinate values, a dumpave file, with the columns needed, and prints the matching line of the metafile that tells ndfes the window's umbrella centre, force constant, temperature and data file. It is usually run in a loop over all windows. (UNVERIFIED) ndfes-PrepareAmberData.py A utility script in the FE-ToolKit library that finds the equilibrated part of an umbrella sampling simulation by forward and reverse analysis of the time series of its biasing potential. It can then write a shorter trajectory or data file that keeps only statistically independent samples from that part, optionally with cpptraj. It is normally run before the data are prepared for free energy analysis with ndfes. (UNVERIFIED) ndfes-CheckEquil.py A utility script in the FE-ToolKit library that merges 1 or more ndfes metafiles, the input files that list each biased simulation with its umbrella centre, force constant and data file, into a new metafile. It is used to analyse the combined sampling of several independent trials or of several iterations of the finite temperature string method. It can also move a metafile to another directory by adjusting the relative paths it lists. (UNVERIFIED) ndfes-CombineMetafiles.py A utility script in the FE-ToolKit library that prints the free energy surface stored in an ndfes checkpoint file as plain text, 1 line per occupied histogram bin. Each line gives the bin centre on each reaction coordinate, the free energy and its standard error or 95% confidence interval, a reweighting entropy that warns when few samples dominate the bin, and the number of samples. The output is used to plot the free energy surface. (UNVERIFIED) ndfes-PrintFES.py A utility script in the FE-ToolKit library that averages free energy surfaces stored in ndfes checkpoint files, such as those from independent repeats of a simulation, and writes the result as a new checkpoint file. Each input surface is shifted to fit the average as closely as possible, and the spread between the repeats is used to give a better uncertainty for the best free energy estimate. (UNVERIFIED) ndfes-AvgFESs.py A program in the FE-ToolKit library that finds and evaluates paths on a multidimensional free energy surface computed by ndfes. It evaluates the free energy along a given curve, optimizes a minimum free energy path between 2 states, or prepares the next set of simulations for the finite temperature string method. A minimum free energy path is the route of lowest free energy connecting 2 stable states, and it shows how a conformational change or reaction proceeds. (UNVERIFIED) ndfes-path A utility script in the FE-ToolKit library that prepares an initial guess for the finite temperature string method from restraint definitions for 2 or more states, such as reactant and product. It writes input and restraint files for a chosen number of windows connected by linear interpolation, and shell scripts that generate a starting structure for each window by a series of short simulations from the nearest prepared structure. The path is then refined by later string iterations. (UNVERIFIED) ndfes-path-prepguess.py A utility script in the FE-ToolKit library that prepares the latest set of simulations of a finite temperature string calculation for analysis. It acts as a wrapper that runs ndfes-CheckEquil.py, ndfes-PrepareAmberData.py and ndfes-CombineMetafiles.py, after which ndfes estimates the free energy surface and ndfes-path updates the path. The finite temperature string method refines a minimum free energy path by repeated rounds of simulation along it. (UNVERIFIED) ndfes-path-analyzesims.py A utility script in AmberTools that computes the reference energy a new titratable residue needs in constant pH or constant redox potential simulations. It titrates the model compound with repeated short constant pH runs, adjusting the reference energy until the residue is 50% protonated at a pH equal to the model compound's pKa. It is an alternative to computing the reference energy by thermodynamic integration. (UNVERIFIED) finddgref.py A utility script that extracts the electrostatic potential and the points it was computed at from quantum chemistry output, and writes them in the form a charge fitting program expects. It is a format conversion between 2 steps of charge derivation. It performs no fitting itself. (UNVERIFIED) espgen.py A pipeline script in AmberTools that sets up and runs restrained electrostatic potential (RESP) charge fits over several conformations and several molecules at once. It generates conformers with antechamber and RDKit, reads the electrostatic potential from Gaussian or GAMESS output or computes it with QUICK, then writes the fitting input with pyresp_gen.py and performs the fit with py_resp.py. It is meant to make the derivation of RESP charges easier and more standardised. (UNVERIFIED) pype-resp.py The chamber script is a utility within the AmberTools suite that facilitates the conversion of CHARMM-formatted force field files into a format that can be used by AMBER. It processes CHARMM topology (PSF) and parameter (PAR) files, allowing users to run simulations with CHARMM force fields using the AMBER simulation engines. chamber match_atomname is a utility program within the AmberTools suite that takes an input file and a reference file in pdb, ac, prepi, prepc and mol2 format and it automatically detects the corresponding atom name in the reference for each atom name in the input. An output file in the same format as that of the input is generated using the matched atom names. Designed to recover atom-name information lost after running Gaussian calculations. match_atomname A software library implementing the message passing standard, providing the functions a parallel program calls to send and receive data between machines. The standard says what the functions do; a library is a working version of them. Which one is installed mainly affects performance. (UNVERIFIED) MPI library An implementation of the message passing standard that moves data directly between graphics processors on different machines, without routing it through their central processors. Avoiding that detour removes a large part of the delay in a simulation running across several machines with accelerators. It requires network hardware that supports direct access. (UNVERIFIED) MVAPICH2-GDR A software library for finding a few eigenvalues and eigenvectors of a very large matrix without forming the whole matrix. Normal mode analysis of a large structure needs exactly that, since only the slowest modes are of interest. Computing every mode would cost far more for a system of that size. (UNVERIFIED) ARPACK An analysis program, the predecessor of cphstats in AMBER, that computed predicted pKa values and fractions of protonation for the titratable residues from the output of a constant pH molecular dynamics run. Its standard output format is kept by cphstats, which replaced it. (UNVERIFIED) calcpka A system preparation program that rewrites a topology file to contain several non-interacting copies of a chosen region of a molecule, as locally enhanced sampling requires. The copies feel the rest of the system but not each other. It is run before the simulation, because the method cannot be switched on afterwards. (UNVERIFIED) addles A formal system of instructions and syntax rules in which computer programs are written. Which one a tool is written in affects how fast it runs and how easily it can be changed or extended. Reproducing a computational result may require the same language version. (UNVERIFIED) programming language NAB, or Nucleic Acid Builder, is a specialized programming language designed specifically for the purpose of constructing and manipulating three-dimensional models of nucleic acids and proteins. For experts, NAB provides a powerful and scriptable way to build complex or non-standard biomolecular structures that would be difficult to create with standard graphical tools. Originally distributed with the AmberTools software suite, it is now only available from an independent GitHub repository. NAB The FE-ToolKit is a collection of tools designed to facilitate the setup, execution, and analysis of alchemical free energy simulations and umbrella sampling. FE-ToolKit An analysis program that computes a theoretical small-angle X-ray scattering profile from a structure together with a solvent distribution from 3D-RISM theory. Including the ordered solvent layer matters because that layer scatters measurably. It differs from the trajectory-based tool in obtaining the solvent from theory rather than from sampling. (UNVERIFIED) saxs_rism An analysis program in AmberTools that computes a theoretical small-angle X-ray scattering profile from snapshots of 2 explicit-solvent simulations: a sample containing the biomolecule with water and ions, and a blank containing the solvent alone. Subtracting the blank and averaging over the sampled conformations makes the result comparable with a measurement on a solution. It differs from saxs_rism in taking the solvent from simulation rather than from 3D-RISM theory. (UNVERIFIED) saxs_md A collection of separate programs distributed and installed together, sharing file formats and conventions so that the output of one is directly usable by the next. AmberTools and GROMACS are examples. The distinction from a single program matters, because a suite's name is often used loosely for the one program in it that a study actually ran. (UNVERIFIED) software suite A software suite in AmberTools that analyses volumetric data, such as 3-dimensional densities of water and ions, produced by simulations, 3D-RISM or Poisson-Boltzmann calculations, or by experiments such as crystallography and cryo-electron microscopy. Its functions, used mainly through the metatwist program, include converting map formats, reducing maps to radial or helical profiles, filtering them, and placing water and ions at density peaks. (UNVERIFIED) MoFT An analysis program in AmberTools, the command-line front end to the MoFT library, that processes 3-dimensional density maps of solvent and ions from simulations, 3D-RISM calculations or experiments. It computes radial distribution functions and twisted or untwisted maps of ion and water density around nucleic acid helices, filters and converts maps, and places explicit water molecules and ions at density peaks. (UNVERIFIED) metatwist A trajectory analysis program, formerly part of the AMBER suite, that read a script of commands and computed properties such as root mean square deviation, distances and angles from a coordinate file. It has been superseded by cpptraj, which is faster and handles more formats. Older published workflows still refer to it. (UNVERIFIED) ptraj An analysis method that converts a continuous solvent density map into a list of discrete water positions, by repeatedly taking the highest remaining peak and removing the density around it. A simulation needs individual water molecules, and a theory-derived map gives only a density. It is the step that turns one into the other. (UNVERIFIED) placevent analysis A simulation engine, distributed with the AMBER package rather than with the free AmberTools, optimised for speed on large systems and on graphics processing units. It supports fewer methods than the general-purpose engine sander but runs the common ones far faster. It is the engine most production molecular dynamics in AMBER is run with. (UNVERIFIED) PMEMD A simulation engine, distributed with AmberTools, that performs energy minimisation, molecular dynamics and some free energy calculations. It is the general-purpose engine of the suite and supports the widest range of methods. It is slower than pmemd, the optimised engine of the separately licensed AMBER package, which is the trade for that breadth. (UNVERIFIED) sander A software service, reached through a web browser, that runs a computational tool or searches a database on the user's behalf and returns the result. Nothing is installed locally, so people can use methods they could not otherwise run. The cost is that the input leaves the user's machine, and the exact software version used may not be recorded. (UNVERIFIED) web server GLYCAM-Web is a collection of online tools and web servers that assist researchers in building three-dimensional models of carbohydrates and glycoproteins. It uses the GLYCAM force field to generate realistic structures and provides tools for tasks like predicting the conformation of glycosidic linkages. GLYCAM-Web The Glycoprotein Builder is a web-based tool that automates the process of constructing three-dimensional models of glycoproteins. It allows users to start with a protein structure and graphically attach various glycan (sugar chain) structures to specific amino acid residues. For researchers, this tool greatly simplifies the otherwise complex and tedious task of building a complete, simulation-ready model of a glycosylated protein. Glycoprotein Builder An open-source software library for working with chemical structures, covering reading and writing file formats, computing molecular descriptors, generating 3-dimensional conformations and comparing molecules. It is the foundation most cheminformatics workflows in Python are built on. It handles the chemistry around a simulation rather than the simulation itself. (UNVERIFIED) RDKit A software library for running molecular dynamics, designed to be called from a program or script rather than run as a command. It is built with graphics processors in mind and is among the fastest engines available on them. Because it is a library that is easy to extend, it is often used to build and test new simulation methods. (UNVERIFIED) OpenMM A standard specifying how parts of a parallel program running on separate machines exchange data by sending messages to one another. Because the machines share no memory, every piece of shared information must be sent explicitly. It is how a single simulation is spread across the nodes of a cluster. (UNVERIFIED) Message Parsing Interface MPI A programming interface that lets parts of a program run on several processor cores that share the same memory, by marking sections of the code rather than rewriting it. It is how most simulation engines use the cores of a single machine. It does not spread work across separate machines, which needs a different mechanism. (UNVERIFIED) OpenMP A software library, and the standard interface it defines, for low-level linear algebra routines such as multiplying a matrix by a vector or by another matrix. Hardware vendors supply versions tuned for their own processors, so the same program runs at very different speeds depending on which is installed. Almost every numerical library is built on top of it. (UNVERIFIED) BLAS A software library providing routines for linear algebra problems such as solving systems of equations, finding eigenvalues and eigenvectors, and least-squares fitting. Normal mode analysis, principal component analysis and many other analyses reduce to calls into it. Its speed determines how large a matrix those analyses can handle. (UNVERIFIED) LAPACK DL-Find is a software library that provides a collection of robust algorithms for geometry optimization tasks. It contains routines for efficiently finding energy minima (stable structures) and first-order saddle points (transition states) on a potential energy surface. This library acts as an optimization engine that can be incorporated into other simulation or quantum chemistry programs to perform complex structural searches and reaction path calculations. DL-Find A software library of AMBER-compatible molecular mechanics routines, written for the NAB molecular manipulation language, that computes energies and forces for minimization and molecular dynamics with an emphasis on implicit solvent models such as generalized Born, Poisson-Boltzmann and RISM. It has been moved out of AmberTools into a separate repository together with NAB and its C interface. (UNVERIFIED) sff x3DNA is a comprehensive software package for the analysis, rebuilding, and visualization of three-dimensional nucleic acid structures. It takes a DNA or RNA structure file as input and calculates a wide range of standard helical and base pair parameters that are used to describe the detailed geometry of the helix. It is a cornerstone tool in the field of nucleic acid structural biology for characterizing and comparing experimental or simulated structures. x3DNA 3DNA A software library that computes energies and forces from data-driven many-body potentials, MB-pol for water and MB-nrg for ions and small molecules, for use by molecular dynamics programs such as LAMMPS and i-PI. These potentials add explicitly fitted terms for pairs and triples of molecules to a polarizable model that covers the remaining many-body effects. Fitting them to high-level quantum calculations gives close to quantum chemical accuracy from small clusters to liquids. (UNVERIFIED) MBX A match analysis tool is a program that compares two or more molecular structures or datasets to find and quantify their similarities. This can involve superimposing two protein structures to calculate their RMSD or searching a database of small molecules to find ones that have a similar shape or chemical features to a query compound. match analysis tool A software framework, part of the AmberTools free energy workflow FEW, that automates the calculation of ligand binding free energies by the MM-PBSA and MM-GBSA methods from snapshots of molecular dynamics simulations. It offers 4 variants of the Poisson-Boltzmann calculation and 3 of the generalized Born calculation, applied either to 1 trajectory of the complex or to separate trajectories of complex, receptor and ligand. (UNVERIFIED) Weighted Averaging Momentum Method WAMM A software framework, part of the AmberTools free energy workflow FEW, that prepares and analyses the simulations needed for a linear interaction energy estimate of ligand binding. It sets up runs of each ligand bound to its receptor and free in solution, and computes from them the ligand's electrostatic and van der Waals interaction energies, treating many ligands with 1 consistent protocol. The energy differences are turned into binding free energies with empirical coefficients, which can be fitted by linear regression against measured data. (UNVERIFIED) Linear Interaction Energy Workflow LIEW A software framework, part of the AmberTools free energy workflow FEW, that automates relative binding free energy calculations between 2 ligands by thermodynamic integration. It sets up the transformation simulations at each coupling state and, once they have run, integrates the results to give the free energy difference. It is the thermodynamic integration counterpart to the linear interaction energy workflow. (UNVERIFIED) Thermodynamic Integration Workflow TIW An algorithm that computes an approximate solution to a mathematical equation or a system of equations that has no exact formula for its solution or is too large to solve directly. Many such solvers improve a trial solution step by step until it is accurate enough. Examples are the conjugate gradient methods used to solve the large systems of linear equations that arise in implicit solvent models such as the Poisson-Boltzmann equation. (UNVERIFIED) numerical solver algorithm An iterative algorithm for solving a large system of linear equations whose matrix is not symmetric. The ordinary conjugate gradient method requires a symmetric matrix and can fail without one. In Poisson-Boltzmann continuum electrostatics it is used with discretization schemes, such as the immersed interface method, that produce a matrix that is not symmetric. (UNVERIFIED) biconjugate gradient solver algorithm A numerical solver algorithm that speeds up an iterative calculation by combining several saved trial solutions, weighted so that the combined error left in the equations, the residual, is as small as possible. Each new trial solution is built from this combination with a set step size, and in AMBER the method accelerates the solution of the 1D-RISM and 3D-RISM integral equations of solvation theory. It is a modified form of the method used to converge self-consistent field calculations in quantum chemistry. (UNVERIFIED) modified direct inversion in the iterative subspace solver algorithm An algorithm that advances the positions and velocities of the atoms from one step to the next, using the forces computed at the current step. It approximates the continuous equations of motion by a series of short, discrete time steps. Integrators differ in accuracy, stability at large time steps, cost, and how well they conserve energy over a long run. (UNVERIFIED) integration algorithm An integration algorithm that updates positions from the current velocity and the forces at the current and previous steps, and updates velocities from the forces at the previous, current and new steps. It produces the same positions as the Verlet algorithm but computes velocities more accurately at each point. That matters when a quantity depending on velocity, such as kinetic energy, is being reported. (UNVERIFIED) Beeman integrator A numerical solver algorithm that solves a large, sparse, symmetric system of linear equations by the conjugate gradient method, first transforming the system with an approximate, deliberately incomplete Cholesky factorisation of its matrix to reduce the number of iterations. An exact factorisation would cost as much as solving the system directly, so only part of it is kept. It is the default solver for the finite-difference Poisson-Boltzmann equation in the pbsa program of AMBER. (UNVERIFIED) modified incomplete Choleski conjugate gradient solver algorithm An algorithm that computes the approximate value of a definite integral from values sampled at discrete points. It is used wherever the function being integrated is known only at the points where it was measured or computed. The free energy of a thermodynamic integration calculation is obtained this way. (UNVERIFIED) numerical integration algorithm A numerical integration algorithm for integrating a function over the surface of a sphere, using a fixed set of points and weights chosen to be exact for spherical harmonics up to a given order. It reaches a given accuracy with far fewer points than an evenly spaced arrangement. It supplies the angular part of the grid used in density functional theory. (UNVERIFIED) Lebedev quadrature algorithm An algorithm that computes a property or finds a pattern in raw simulation data. It is the concrete procedure an analysis method is carried out by, as distinct from the idea of the method itself. Two programs implementing the same method can use different algorithms and give slightly different numbers. (UNVERIFIED) analysis algorithm An analysis algorithm that examines or transforms data recorded as a sequence in time. A trajectory is such a sequence, so techniques built for signals apply to it directly. Examples are smoothing, filtering and splitting a series into its component frequencies. (UNVERIFIED) signal processing algorithm A signal processing algorithm that removes frequencies above or below a chosen cut-off frequency while passing the remaining frequencies with a response that is as flat, that is as free of ripples, as possible. Flatness in the kept range is what it optimises, at the cost of a gradual rather than sharp cut-off. It is used to separate slow collective motion in a trajectory from fast thermal noise. (UNVERIFIED) butterworth filter algorithm A numerical integration algorithm that approximates the area under a curve by joining neighbouring points with straight lines and summing the resulting trapezoids. It needs only the sampled values and treats the function as a straight line between each pair of neighbouring points. It is a common way to integrate the thermodynamic integration curve over the coupling parameter. (UNVERIFIED) trapezoid integration algorithm An algorithm designed to produce a sequence of numbers that shows no discernible pattern and so appears to be random. In computational science, such generators supply the element of chance that stochastic simulation methods need. In simulation these are pseudo-random generators, which compute the sequence from a starting value called a seed, and they supply the random numbers used by stochastic thermostats and Monte Carlo methods. (UNVERIFIED) random number generator algorithm A pseudo-random number generator of the family developed by Marsaglia, built from simple shift and multiply operations on a small internal state. The simplicity makes them fast and easy to implement, which is why they appear inside simulation engines. They pass standard statistical tests while using far less state than a long-period generator. (UNVERIFIED) Marsaglia random number generator algorithm A pseudo-random number generator, based on a twisted generalized feedback shift register, whose sequence repeats only after a period of 2^19937 - 1 in its standard MT19937 form. It is the default generator in many scientific libraries. It is not suitable where an adversary must not predict the sequence, which does not arise in simulation. (UNVERIFIED) Mersenne Twister random number generator algorithm A pseudo-random number generator that advances a linear congruential generator and passes its state through an output function, such as an xor-shift followed by a rotation whose size is taken from the state, to remove the patterns of the plain congruential sequence. The result has much better statistical quality than a linear congruential generator while staying small and fast. (UNVERIFIED) permuted congruential random number generator algorithm A pseudo-random number generator that updates 128 bits of state with xor, shift and rotate operations and produces each output by adding and rotating parts of the state. It is fast and passes current statistical test suites, and its small state makes it cheap to keep one per thread. Giving each thread its own generator and seed helps keep a parallel simulation reproducible. (UNVERIFIED) Xoshiro128++ random number generator algorithm MDposit is a web-based platform and portal designed for the sharing, visualization, and analysis of molecular dynamics simulation data. It serves as a distributed database interface that allows users to browse, access, and interactively view trajectories from different research groups. This tool aims to facilitate the dissemination and reuse of large-scale simulation datasets within the scientific community. MDposit An algorithm that computes the discrete Fourier transform of N values, converting data between a description in time or space and one in frequency, in a time that grows as N log N. Computing the same transform directly would cost far more, growing with the square of N. It makes particle mesh Ewald affordable, because that method needs a forward and an inverse transform of the charge grid at every step. (UNVERIFIED) fast fourier transform algorithm A clustering algorithm that grows a group outward from any point with enough neighbours within a chosen distance, and leaves sparse points unassigned. It needs no target number of groups and can find groups of any shape. Structures left unassigned are treated as rare conformations rather than forced into a group, which suits a trajectory well. (UNVERIFIED) DBSCAN clustering algorithm A clustering algorithm that divides structures into a fixed number of groups by repeatedly assigning each to the nearest group centre and then recomputing the centres. The number of groups must be chosen in advance, which is its main limitation. It also assumes groups are roughly round and of similar size, which conformations often are not. (UNVERIFIED) k-means clustering algorithm A clustering algorithm that picks group centres as points which have many close neighbours and are also far from any point with still more neighbours. Those 2 conditions identify a peak in the crowding rather than an arbitrary starting point. Every other point then joins the nearest denser point, which assigns the whole set in one pass. (UNVERIFIED) density peaks clustering algorithm An optimization algorithm for non-linear least-squares fitting of a model to data that blends 2 methods: cautious gradient descent steps when far from the best fit and fast Gauss-Newton steps when close to it. Shifting between them according to how well the fit is going makes it converge reliably even from a poor starting guess. It is often used to fit force field parameters to experimental or quantum mechanical reference data. (UNVERIFIED) Levenberg-Marquardt algorithm A force field component that holds the complete definition of each molecular building block, or residue, giving its atoms, their types and charges, and the bonds between them, usually stored as a data file. It lets a topology be assembled from a sequence without every atom being specified. A residue absent from it cannot be built automatically. (UNVERIFIED) residue library An optimization algorithm that keeps a set of trial points, 1 more than the number of variables, forming a simplex, and repeatedly replaces the worst point by reflecting it through the center of the others, expanding or shrinking the set as needed. It never needs a derivative, only the ability to evaluate the function. That makes it usable where derivatives are unavailable or unreliable, at the cost of converging slowly. (UNVERIFIED) downhill simplex minimizer algorithm A set of effective atomic radii fitted for use with the SMx family of implicit solvent models. The radii absorb part of the error of the underlying model, which is why they are specific to it. Substituting a general-purpose set such as Bondi changes the computed solvation energies. (UNVERIFIED) Hawkins-Cramer-Truhlar radii set A numerical integration algorithm that evaluates a function at specially chosen points, with matching weights, so the result is exact for polynomials up to a given degree. Choosing the positions as well as the weights doubles the degree that a given number of points can handle. It needs the function to be evaluable anywhere, not only at points already sampled. (UNVERIFIED) Gaussian quadrature algorithm An algorithm that computes the area of a molecule's surface that a solvent molecule can reach, called the solvent accessible surface area. Numerical methods, such as the Shrake-Rupley method, place test points on a sphere around each atom enlarged by the radius of a solvent probe and count those that no other atom covers, while analytical methods, such as LCPO, estimate the area from how much the atom spheres overlap. The result is used to estimate the nonpolar part of the solvation free energy in implicit solvent models and to measure how exposed parts of a molecule are. (UNVERIFIED) surface area calculation algorithm An algorithm that estimates solvent-accessible surface area from a sum over overlapping pairs of atoms, with fitted coefficients, rather than by placing test points. Having a closed form makes it fast enough to run at every step of a simulation, and to be differentiated so it contributes forces. Point-based methods can be more accurate but are slower, and counting points does not give smooth forces. (UNVERIFIED) LCPO algorithm An algorithm that measures how far 2 structures differ, by first superposing them to remove any overall shift or rotation and then taking the root mean square of the remaining distances between matched atoms. Without the superposition step, moving a molecule across the box would register as a large change in shape. The choice of which atoms to superpose on decides what the number means. (UNVERIFIED) RMSD calculation algorithm An algorithm that finds the optimal superposing rotation as the eigenvector of the largest eigenvalue of a 4-by-4 symmetric matrix built from the 2 sets of points, instead of by decomposing a 3-by-3 matrix as the Kabsch algorithm does. The eigenvector is a unit quaternion, a set of 4 numbers that always describes a proper rotation, so the result can never be a reflection, which matrix methods must check for separately. It gives the same optimal superposition as the Kabsch algorithm. (UNVERIFIED) quaternion RMSD algorithm An optimization algorithm that builds an approximation to the curvature of a function from the slopes seen at successive steps, and uses it to choose each new step. Approximating the curvature rather than computing it converges in far fewer steps than using slope alone, without the cost of second derivatives. It stores a full matrix, which limits it to problems with a moderate number of variables. (UNVERIFIED) BFGS BFGS algorithm Broyden-Fletcher-Goldfarb-Shanno algorithm A minimization algorithm that chooses each step from the current downhill direction combined with the previous step's direction, so that it does not undo progress already made. That memory is what lets it follow a long narrow valley, where steepest descent zig-zags. It is the usual choice once the worst overlaps in a structure have been removed. (UNVERIFIED) conjugate gradient algorithm An optimization algorithm that finds a minimum by searching along one direction at a time and gradually replacing those directions with better ones built from the progress made. It never needs a derivative, only function values. It suits problems where derivatives are unavailable or too noisy to trust. (UNVERIFIED) Powell's algorithm An integration algorithm that updates different forces at different intervals, evaluating fast-changing forces, such as bonded forces, every step and slow-changing forces, such as long-range forces, only every few steps. When the slow forces are the expensive ones, evaluating them less often saves much computing time. Making the outer interval too long makes the simulation unstable, because the slow-force updates can fall into resonance with the fastest motions. (UNVERIFIED) multiple time-stepping algorithm A multiple time-stepping algorithm that splits the forces into groups that change quickly or slowly and nests their updates, integrating the fast forces with a short inner timestep and applying the slow forces as impulses at a longer outer timestep. It is derived so that the scheme stays time-reversible, and it conserves energy well as long as the outer timestep is short enough, about 4 fs in typical biomolecular simulations, to avoid resonance with the fastest motions. (UNVERIFIED) r-RESPA integrator An algorithm that computes long-range forces quickly by grouping distant particles into a hierarchy of cells and treating each distant group as a single source. Seen from far enough away, a group acts almost like a single point, so computing the effect of each member separately is wasted work. It reduces the cost from growing with the square of the particle count to growing close to linearly. (UNVERIFIED) Treecode algorithm A multiple time-stepping algorithm that estimates an expensive force, such as the 3D-RISM solvation force, at intermediate steps from forces already computed at earlier steps instead of computing it in full. The current coordinates are written as a weighted combination of the coordinates at earlier fully computed steps, and the same weights are applied to the forces from those steps. This allows long intervals between full calculations, but unlike r-RESPA the scheme does not conserve energy and is not time-reversible. (UNVERIFIED) force-coordinate extrapolation integrator An algorithm that decides, during a simulation, which atoms belong in the quantum region at each moment. It is needed when molecules such as solvent move into and out of the region of interest, because a fixed list would keep treating molecules that have left as quantum and miss molecules that have arrived. The difficulty is changing the assignment without a jump in the energy or forces. (UNVERIFIED) QM/MM region selection algorithm A region selection algorithm that uses 2 distance thresholds, a nearer one for entering the quantum region and a farther one for leaving it. The gap between them stops an atom sitting near the boundary from switching back and forth every step. With a single threshold, atoms that cross and re-cross the boundary make the region definition change rapidly. (UNVERIFIED) hysteretic region selection algorithm An algorithm that approximately solves the equations of quantum mechanics for a molecular system, to obtain how its electrons are distributed and the energy of that arrangement. The equations can be solved exactly in closed form only for systems with a single electron, so practical methods each make an approximation. Methods differ in which approximation they make and in how much computing time they need. (UNVERIFIED) quantum mechanics algorithm An algorithm that makes a quantum calculation on a large molecule possible by splitting it into overlapping pieces, solving each, and combining the results. The cost of a conventional calculation grows much faster than the size of the system, while the cost of this approach grows roughly in proportion to it. The pieces must overlap, or the boundaries between them introduce errors. (UNVERIFIED) divide and conquer algorithm A multiple time-stepping algorithm built on the velocity Verlet integrator in which slowly varying forces are computed only at major steps and applied as impulses, while fast-varying forces are recomputed at every minor step in between. In the mdgx program of AMBER, bond, angle and 1-4 non-bonded terms are recomputed at every minor step, and the other non-bonded and dihedral terms only at major steps. (UNVERIFIED) I/r-RESPA integrator A thermostat algorithm that couples the system to a thermostat variable, that variable to a second one, and so on along a chain, extending the Nosé-Hoover thermostat. The chain corrects a failure of the single-variable form, which may not reach the canonical distribution for small or stiff systems, such as a harmonic oscillator, because their dynamics do not explore all accessible states. (UNVERIFIED) Nose-Hoover chain thermostat A thermostat algorithm that rescales the velocities towards the target temperature by a factor that includes a random part, chosen so that the kinetic energy follows the canonical distribution instead of being driven to a single value. It reaches the target temperature as quickly as simple rescaling schemes such as the Berendsen thermostat, while sampling the canonical ensemble correctly, which they do not. (UNVERIFIED) ntt=11 Bussi thermostat ntt=11 An integration algorithm of the velocity-Verlet family in which the thermostat acts in the middle of each step, between 2 half-step updates of the coordinates, rather than at the start or end of the step. Placing it there makes the sampled distribution of coordinates exact for a harmonic system and much more accurate in general, so a larger timestep can be used for the same accuracy in structural properties. It can be combined with thermostats such as the Langevin and Andersen thermostats. (UNVERIFIED) velocity-Verlet middle integrator A polarization algorithm that gives each induced dipole a small fictitious mass and lets it move under its own equation of motion, instead of solving for it exactly at every step. If the mass is small enough the dipoles stay close to the exact solution while never being solved for. It removes the iteration that otherwise dominates the cost. (UNVERIFIED) Car-Parrinello induced dipole scheme A thermostat algorithm in which the thermostat acts for a full time step in the middle of each integration step, between 2 half-step updates of the coordinates, rather than at the beginning or end of the step. This placement keeps the sampled distribution of coordinates accurate at much larger timesteps, which makes the sampling of structural and thermodynamic properties more efficient. It works with several thermostats, such as the Langevin and Andersen thermostats, and can be combined with Monte Carlo or stochastic cell rescaling barostats. (UNVERIFIED) middle thermostat scheme A set of atomic radii for Poisson-Boltzmann calculations, assigned by AMBER atom type and charge and optimized by Tan and Luo for the standard residues of the AMBER force fields. They define the dielectric boundary between solute and solvent and are the default radii in the AMBER Poisson-Boltzmann program. Molecules built with general AMBER force field atom types do not use them and take their radii from the topology instead. (UNVERIFIED) Tan&Luo radii set A conjugate gradient algorithm that weights the previous search direction by the dot product of the new gradient with the change in gradient since the last step, divided by the squared length of the old gradient. The Fletcher-Reeves variant instead uses only the ratio of the squared lengths of the new and old gradients. When progress stalls and the gradient barely changes, the weight falls towards 0, so the search restarts by itself along the steepest descent direction. (UNVERIFIED) Polak-Ribiere conjugate gradient algorithm A quasi-Newton optimization algorithm, a variant of the BFGS algorithm, that approximates the curvature of a function from the last few steps taken instead of storing the full curvature matrix. Storing that matrix is not practical when there are hundreds of thousands of variables, as in a large molecular system. Keeping only recent history gives most of the benefit at a fraction of the memory. (UNVERIFIED) L-BFGS algorithm L-BGFS minimizer limited-memory BFGS algorithm A minimization algorithm that chooses each step by approximately solving the Newton equations, which use the curvature of the energy surface, with an inner linear conjugate gradient calculation that is stopped early. Using curvature as well as slope needs far fewer steps than slope alone, and the inner calculation needs only products of the curvature matrix with vectors, so the full matrix of second derivatives never has to be stored. Stopping the inner calculation early saves time while the step is still far from the minimum. (UNVERIFIED) truncated Newton conjugate gradient algorithm An enhanced sampling algorithm is a computational procedure that modifies a standard simulation to accelerate the exploration of a molecule's possible states or to observe rare events. These algorithms are designed to overcome the high energy barriers that can trap a simulation in a single state for a long time. enhanced sampling algorithm An enhanced sampling algorithm where multiple independent simulation replicas (walkers) simultaneously explore the conformational space and contribute to a single, shared biasing potential. This parallelization significantly accelerates the convergence of methods like metadynamics or adaptive biasing force. multiple-walkers algorithm The IUPAC International Chemical Identifier (InChI) is a standardized text string that provides a unique, machine-readable signature for a specific chemical substance. It is algorithmically generated from a molecule's structural information and it is sensitive to its chirality, ensuring that a given compound will always produce the same identifier regardless of how it is drawn. This code serves as a universal label for linking chemical information across different databases and software. InChI code The ICOSA algorithm is a computational method used to rapidly calculate the solvent-accessible surface area (SASA) of a molecule. It works by placing a sphere of dots, arranged in a pattern based on an icosahedron, around each atom and then counting how many of these dots are not covered by neighboring atoms. This numerical approach provides a fast and efficient way to estimate the molecular surface area, a property that is crucial for implicit solvent models and for analyzing a molecule's exposure to its environment. ICOSA algorithm A minimization algorithm that moves the atoms at each step in the direction of the force on them, which is the steepest downhill direction from where they are. It is robust, because a small enough step in that direction always lowers the energy. It becomes very slow in a long narrow valley, where the steepest direction points across the valley rather than along it. (UNVERIFIED) steepest descent algorithm An algorithm that identifies which parts of a protein form regular repeating shapes, such as helices and sheets, from the coordinates of its backbone. It works by finding the pattern of hydrogen bonds between backbone groups, or by examining backbone angles. It is what turns a list of coordinates into the helix-and-arrow picture of a protein. (UNVERIFIED) secondary structure assignment algorithm The DSSP (Define Secondary Structure of Proteins) algorithm is the most widely used and standard algorithm for assigning secondary structure to the residues of a protein. It works by identifying the characteristic hydrogen-bonding patterns between the backbone amide and carbonyl groups that define alpha-helices and beta-sheets. The detailed, per-residue output of the DSSP algorithm is a standard tool for protein structure analysis and validation. DSSP algorithm OBSOLETE. A trajectory analysis algorithm is a computational procedure that processes the time-series of atomic coordinates from a simulation to calculate a specific property. These algorithms are the engines of analysis tools, turning raw simulation output into interpretable scientific data. For experts, this includes the specific algorithms for tasks like imaging periodic trajectories, calculating correlation functions, or clustering conformations. candidate for removal per https://github.com/CPCLab/molsim-ontology/issues/21 obsolete trajectory analysis algorithm true An analysis algorithm that unwraps the trajectories of molecules from a constant-pressure simulation with periodic boundary conditions, by adding to each molecule's unwrapped position its displacement from the previous frame, measured in Cartesian space. It removes the jumps that appear when a molecule crosses the box boundary and is put back into the central box, while handling changes in box size between frames. It is used mainly to calculate diffusion coefficients from mean-squared displacements. (UNVERIFIED) toroidal-view-preserving algorithm An algorithm that draws a smooth curve through a set of points by fitting a cubic polynomial between each neighbouring pair and matching their slopes and curvatures where they join. The matching is what stops the curve having visible corners. It is often used to draw a smooth free energy profile from values computed at discrete points. (UNVERIFIED) cubic spline interpolation algorithm A signal processing algorithm that decomposes a signal into contributions at each frequency and each moment at once, by matching it against a short wave shape, the wavelet, that is stretched and shifted across the data. A plain frequency analysis, such as a Fourier transform, does not show when each frequency occurs. That matters for a trajectory, where a motion may appear only during part of the run. (UNVERIFIED) continuous wavelet transform An algorithm that applies the methods of statistics to summarise data, estimate a quantity, or judge how reliable a result is. Simulation output is a finite sample, so averages, probability distributions and their uncertainties all have to be estimated from it. Examples are fitting a probability distribution to sampled values and estimating the error bar on an average. (UNVERIFIED) statistical algorithm A statistical algorithm that estimates the distribution behind a set of points by placing a small smooth peak, called a kernel, such as a Gaussian, on each point and adding them together. It assumes no particular shape for the distribution, unlike fitting a named family. How wide the kernels are decides whether it reveals real structure or smooths it away. (UNVERIFIED) kernel density estimation algorithm A residue name is a standardized short code, typically three or four letters, used in structure files to identify a specific type of molecular building block. For example, 'ALA' is the residue name for the amino acid alanine, and 'DG' can represent a deoxyguanosine nucleotide. residue name A short code identifying a sugar residue in the GLYCAM carbohydrate force field, encoding which sugar it is, its ring form, its anomeric configuration, its D or L form, and which of its positions are linked to other residues. Sugars branch, so the same sugar appears in many linkage patterns and each needs its own template. That is why the code carries linkage information where amino acid codes do not. (UNVERIFIED) GLYCAM residue name An atom name is a unique label given to an atom within a specific residue to distinguish it from other atoms of the same element. For example, in an amino acid, the names 'CA' and 'CB' are used to differentiate the alpha-carbon from the beta-carbon. atom name The atom of a residue designated to form the bond joining it to the neighbouring residue in a chain. Building a polymer from residue templates requires knowing which atom on each side makes that bond. Without it, templates could be listed in order and never joined. (UNVERIFIED) residue connection atom The atom of a residue template that bonds to the preceding residue in a chain. In a protein it is the backbone nitrogen. Together with the tail atom it defines the direction in which templates are joined. (UNVERIFIED) unit head atom The atom of a residue template that bonds to the following residue in a chain. In a protein it is the backbone carbonyl carbon. A template with no tail atom defined cannot be followed by another, which is how a chain terminus is expressed. (UNVERIFIED) unit tail atom OBSOLETE. Miller indices are a set of three integers (h, k, l) that uniquely describe the orientation of a plane of atoms within a crystal lattice. These indices are determined from the positions of the spots in an X-ray diffraction pattern and are fundamental to crystallography. They are used to define the specific crystal faces and to calculate the theoretical diffraction pattern from a structural model. Marked for removal. See: https://github.com/CPCLab/molsim-ontology/issues/26 obsolete Miller indices true A classification recorded in a topology stating what general kind of building block a residue is, such as protein, nucleic acid, water, ion or solvent. Analysis tools use it to select parts of a system without naming every residue. A mis-set value makes a selection quietly omit or include the wrong atoms. (UNVERIFIED) residue type A constituent part of a 3-dimensional data map, being either a grid point or the value stored at it. The grid points fix where the map is sampled and the values say what was found there. (UNVERIFIED) volumetric data component A single cell of a 3-dimensional grid, holding one value for the small volume of space it covers. It is to a volume what a pixel is to an image. Analyses that map a property through space, such as where water sits around a protein, report their results one of these at a time. (UNVERIFIED) voxel A classification assigned to an atom to describe its fundamental chemical identity or its specific physical and computational properties within a molecular model. Previous definition: An atom type is a classification assigned to an atom in a force field that defines its interaction parameters based on its chemical environment. It encodes not just the chemical element but also its hybridization state and bonding pattern, so a carbonyl carbon will have a different atom type than an aromatic carbon. atom type A description of an atom's surroundings used by a machine-learned force field, computed from the positions of its neighbours rather than assigned from a table. A conventional atom type is a fixed label such as "aromatic carbon"; this instead varies continuously with the local geometry. That is what lets such a model describe environments it was never given a label for. (UNVERIFIED) neural network atom type A code identifying an amino acid that has been chemically altered, such as by phosphorylation or methylation. The standard 3-letter codes cover only the 20 unmodified residues, so a modified one needs a code of its own. The code is what links the residue to the extra parameters its modification requires. (UNVERIFIED) modified amino acid residue name A path image is a single, discrete snapshot or structure that represents one point along a continuous pathway between two molecular states. A collection of these images is used in methods like the Nudged Elastic Band (NEB) to map out a reaction path or conformational transition. path image Trajectory post-processing is the general process of cleaning and preparing raw simulation trajectory files before performing detailed scientific analysis. This typically involves essential steps like removing the overall rotation and translation of a target molecule to keep it centered, and fixing any molecules that appear "broken" by the periodic boundary conditions. These routine procedures are necessary to create a clean trajectory that is suitable for correct visualization and accurate calculation of structural properties. trajectory post-processing A system preparation stage in which a correspondence is established between the atoms of 2 structures, saying which atom of one corresponds to which atom of the other. An alchemical transformation needs it to know which atoms change and which stay the same. A wrong correspondence produces a calculation that runs to completion and gives a wrong answer. (UNVERIFIED) match list preparation A preparation process that produces the 2 files that describe the system for a simulation: a topology file saying what the system is made of, and a coordinate file saying where every atom is at the start. The topology lists the atoms, the bonds between them and the force field parameters, and it does not change during the run. The coordinates change at every step, which is why the 2 are kept apart. (UNVERIFIED) topology and coordinate file preparation A kinetic property describing the random tumbling of a molecule in a liquid, caused by collisions with the solvent around it, usually expressed as a rotational diffusion coefficient or a rotational correlation time. How fast the molecule tumbles depends on its size and shape and on how viscous the solvent is. NMR relaxation measurements are sensitive to it, which is how a simulated value can be checked. (UNVERIFIED) rotational diffusion A modelling method that rebuilds an all-atom structure from a coarse-grained one, placing the atoms each grouped site stands for. Information was discarded when the coarse model was built, so the rebuilt atoms are a plausible arrangement rather than a recovered one. It is used to inspect a coarse-grained result in atomic detail, or to continue a simulation at finer resolution. (UNVERIFIED) backmapping modeling NMR refinement is a computational process used to generate or improve a three-dimensional model of a molecule using experimental data from Nuclear Magnetic Resonance (NMR) spectroscopy. The method involves finding a structure, or an ensemble of structures, that is both stereochemically realistic and consistent with a set of experimental restraints, such as distances between atoms derived from the Nuclear Overhauser Effect. This process is essential for translating the raw data from an NMR experiment into a high-resolution atomic model. NMR refinement A simulation preparation stage in which the LEaP program of AMBER, run as tleap or xleap, assembles a molecular system and assigns force field parameters, producing the topology and coordinate files an AMBER simulation needs. It covers loading the structure, choosing the force field, adding solvent and counter-ions, and writing the output. It is the step at which a structure becomes a system that can be simulated. (UNVERIFIED) LEaP preparation A procedure applied at intervals during a simulation that sets the total linear momentum of the system to zero, and in some cases also the total angular momentum, so that the system as a whole does not translate or rotate. Small numerical errors, for example in the long-range forces, can set the whole system slowly moving, and the kinetic energy of that overall motion would be counted wrongly as thermal motion and raise the reported temperature. Removing it keeps the reported temperature a measure of the internal motion only. (UNVERIFIED) momentum removal A chemical change made to a protein after it has been built from its amino acid sequence, such as attaching a phosphate, a sugar or a lipid. It can switch the protein's activity on or off, mark it for destruction, or send it to a particular part of the cell. Standard force fields do not include the modified residues, so extra parameters are needed to simulate one. (UNVERIFIED) post-translational modification A post-translational modification in which a branched sugar chain is attached to the nitrogen of an asparagine side chain. It is among the most common modifications of proteins on the cell surface and of secreted proteins. The sugar chains are large, flexible and highly variable, which makes them among the hardest parts of a protein to simulate. (UNVERIFIED) N-linked glycosylation A post-translational modification in which sugars are attached to the oxygen of a side chain, most often of serine or threonine, and are added 1 at a time rather than as a preassembled block. The resulting chains range from a single sugar to long chains, as in the mucins. Their size and variability make such glycosylated proteins hard to model and simulate. (UNVERIFIED) O-linked glycosylation A model specifying the mathematical form assumed for the many-electron wavefunction in a quantum chemistry calculation, such as a single determinant or a combination of several configurations. It fixes what electronic structure the calculation is able to represent before any basis set is chosen. A single determinant cannot describe bond breaking or a near-degenerate ground state. (UNVERIFIED) wavefunction model A wavefunction model in which every occupied molecular orbital holds exactly 2 electrons of opposite spin. It applies to molecules with no unpaired electrons, which is most stable organic molecules in their ground state. It is the default assumption of a quantum chemistry calculation unless an open-shell model is asked for. (UNVERIFIED) restricted closed-shell wavefunction A wavefunction model in which electrons of opposite spin may occupy different spatial orbitals, used mainly for molecules with unpaired electrons such as radicals and many metal complexes. This extra freedom describes unpaired electrons well, but the result can mix in other spin states, which is called spin contamination. It is the basis of unrestricted Hartree-Fock and unrestricted density functional theory calculations. (UNVERIFIED) unrestricted open-shell wavefunction A region of a polypeptide chain classified by the local shape of its backbone, which in the regular forms is held in place by hydrogen bonds between nearby backbone groups. The regular forms are helices and beta strands, and the irregular stretches between them are classed as turns, bends or coil. One chain contains many such regions. (UNVERIFIED) secondary structure A beta sheet is a common shape in proteins where parts of the amino acid chain line up side-by-side like ribbons in a flat, pleated sheet. It is a secondary structure motif formed by two or more extended polypeptide chains, called beta-strands, which are linked laterally by a regular pattern of hydrogen bonds between their backbone atoms. These sheet structures form the rigid core of many proteins and are critical for creating binding surfaces for other molecules. (UNVERIFIED) beta sheet A parallel beta sheet is a type of beta sheet where the adjacent protein chains all run in the same direction, like lanes on a one-way highway. In this arrangement, the neighboring beta-strands are all aligned in the same N-terminus to C-terminus direction, which results in a pattern of angled, evenly spaced hydrogen bonds between the strands. This arrangement is typically found buried in the core of a protein and often requires long loop connections between the strands. parallel beta sheet An anti-parallel beta sheet is a type of beta sheet where the adjacent protein strands run in opposite directions, like traffic on a two-way street. In this arrangement, the neighboring beta-strands are aligned in opposite N-terminus to C-terminus directions, resulting in a pattern of straight, perpendicular hydrogen bonds between the strands. This arrangement is generally more stable than the parallel form and is a very common structural motif found in many proteins. anti-parallel beta sheet OBSOLETE. A helix is a common shape in proteins and DNA where the molecular chain is twisted into a spiral or corkscrew. It is a secondary structure motif characterized by a repeating pattern of backbone dihedral angles that causes the polymer chain to follow a helical path, which is stabilized by a regular pattern of internal hydrogen bonds. The alpha-helix in proteins and the double helix of DNA are two of the most fundamental and recognizable structures in biology. Retired 2026-08-05 in favour of the Sequence Ontology term, which covers the same concept in the ontology whose scope owns protein secondary structure. The MOLSIM hierarchy position is preserved: the SO class is asserted under MOLSIM:001781 secondary structure so the DSSP category set stays in one place. obsolete helix true OBSOLETE. The 3-10 helix is a type of spiral shape found in proteins that is tighter and more narrow than the common alpha-helix. It is a secondary structure motif characterized by a repeating hydrogen bond between the backbone carbonyl of a residue and the backbone amide of the residue three positions down the chain (i → i+3). While less common than the alpha-helix, short segments of 3-10 helix are often found at the ends of alpha-helices, acting as a cap to the structure. Retired 2026-08-05 in favour of the Sequence Ontology term, which covers the same concept in the ontology whose scope owns protein secondary structure. The MOLSIM hierarchy position is preserved: the SO class is asserted under MOLSIM:001781 secondary structure so the DSSP category set stays in one place. obsolete 3-10 helix true OBSOLETE. The alpha-helix is the most common spiral shape found in proteins, resembling a coiled ribbon. It is a right-handed secondary structure motif characterized by a repeating hydrogen bond between the backbone carbonyl of a residue and the backbone amide of the residue four positions down the chain (i → i+4). This highly stable and common structure forms a fundamental building block for the architecture of most proteins. Retired 2026-08-05 in favour of the Sequence Ontology term, which covers the same concept in the ontology whose scope owns protein secondary structure. The MOLSIM hierarchy position is preserved: the SO class is asserted under MOLSIM:001781 secondary structure so the DSSP category set stays in one place. obsolete alpha helix true OBSOLETE. The pi-helix is a wide spiral shape found in proteins that is less stable than the common alpha-helix. It is a secondary structure motif characterized by a repeating hydrogen bond between the backbone carbonyl of a residue and the backbone amide of the residue five positions down the chain (i → i+5). Because its internal packing is less favorable, the pi-helix is usually found only as a short, single-turn insertion within other helical structures. Once thought to be rare, short π-helices are found in 15% of known protein structures and are believed to be an evolutionary adaptation derived by the insertion of a single amino acid into an α-helix. Retired 2026-08-05 in favour of the Sequence Ontology term, which covers the same concept in the ontology whose scope owns protein secondary structure. The MOLSIM hierarchy position is preserved: the SO class is asserted under MOLSIM:001781 secondary structure so the DSSP category set stays in one place. obsolete pi helix true A secondary structure element in which a short stretch of 3 to 5 residues reverses the direction of a protein chain, usually held by a hydrogen bond between backbone groups near its 2 ends. Turns let 2 stretches of chain run back alongside each other, which helps a chain fold into a compact shape. (UNVERIFIED) turn A bend is a region in a protein chain that creates a sharp turn, changing the direction of the polypeptide. It is a short structural motif, typically involving a few amino acid residues, that is characterized by a specific pattern of backbone dihedral angles that reverses the direction of the chain. bend A secondary structure element in which a stretch of polypeptide chain is drawn out into a nearly straight zig-zag, with backbone dihedral angles in the extended region of the Ramachandran plot. It is the conformation of a beta strand, and strands in this conformation pair up through backbone hydrogen bonds to form a beta sheet. The zig-zag places successive side chains on alternating sides of the strand. (UNVERIFIED) extended beta conformation A beta bridge is an isolated, localized hydrogen bond between two amino acid residues from different segments of a protein, mimicking the interactions found in β-sheets. beta bridge A conformational transition is the process where a molecule changes from one stable three-dimensional molecular arrangement to another. This can be a large-scale event, such as a protein switching between an active and an inactive state, or a more local change like the flipping of a sugar ring. Studying these transitions is fundamental to understanding how molecular function is controlled by changes in structure. conformational transition A crankshaft motion refers to a specific, localized, correlated movement of a protein's backbone. Specifically, it is an oscillation or rotation of a rigid peptide plane along an axis defined by two adjacent Cα carbon atoms. crankshaft motion The Solvent MoDeL (MDL) file is an explicit molecular model input used for 1D-RISM and 3D-RISM (Reference Interaction Site Model) calculations in software like Amber. It supplies atomic coordinates, partial charges, and Lennard-Jones parameters for solvent molecules or ions. MDL solvent model format In computational chemistry and molecular dynamics, the mean solvation force is the average net force exerted by solvent molecules on a solute molecule, with all solvent degrees of freedom thermally averaged out. It represents the derivative of the potential of mean force (PMF) and drives conformational changes and structural binding in liquids. mean solvation force The atom position is the precise location of an atom in three-dimensional space. It is described by a set of three Cartesian coordinates (X, Y, Z) that specify its location relative to a defined origin. The collection of all atom positions for a molecule at a given instant constitutes a single "frame" of a simulation trajectory, and the evolution of these positions over time is the fundamental output of a molecular dynamics simulation. atom position An interaction energy component arising when a small amount of electron density moves from one molecule to another at close range. It is a quantum mechanical effect that a force field with fixed atomic charges cannot represent, yet it contributes measurably to hydrogen bonds and to interactions with metal ions. Capturing it explicitly needs a quantum mechanical treatment or a force field with a dedicated charge transfer term. (UNVERIFIED) charge transfer energy An energy chosen as the zero point against which other energies in a calculation are reported. In a free energy calculation it is usually a known or simplified state, and every other state is quoted as a difference from it. Comparisons between systems are meaningless unless the same choice was made for both. (UNVERIFIED) reference energy A physical property is a measurable characteristic of a substance or system that describes its physical state, such as its temperature, pressure, or density. In molecular simulations, these properties are calculated as averages over the course of the trajectory to describe the macroscopic state of the simulated system. These calculated properties are the primary outputs used to validate a simulation against experimental data and to understand the material characteristics of the molecular model. physical property A physical property giving how strongly a single atom scatters X-rays, as a function of scattering angle. It rises with the number of electrons the atom has and falls as the angle increases, because the electron cloud has a finite size. It is the per-atom term from which structure factors are built. (UNVERIFIED) atomic scattering factor A program written in a language that is interpreted as it runs rather than compiled first, used to drive a sequence of steps automatically. Being interpreted makes it quick to change and slow to execute, which suits work that is mostly waiting on other programs. Most of the workflow around a simulation is scripts rather than compiled code. (UNVERIFIED) script An energy arising from the forces between 2 or more separate molecules, or between distinct parts of 1 molecule, computed as the energy of the whole minus the sum of the energies of the parts held at the same geometry. It gathers the electrostatic and van der Waals contributions into a single number. It measures how strongly the parts are held together, but it is not a free energy, because it contains no entropy. (UNVERIFIED) interaction energy An interaction energy between a solute and all the water molecules around it. It sums the electrostatic and van der Waals contributions over every solute-water pair. Following it through a trajectory shows how the solute's contact with water changes as it moves. (UNVERIFIED) solute-water interaction energy An interaction energy summing the interactions of every water molecule with every other water molecule in a system. It measures how strongly the solvent holds itself together, separately from anything the solute does. Changes in it show how a solute disturbs the interactions between the water molecules around it. (UNVERIFIED) water-water interaction energy The REMD round-trip time is a diagnostic metric that measures how long it takes for a single simulation copy to travel through the full range of temperatures in a Replica Exchange Molecular Dynamics simulation. In this method, multiple simulations are run at different temperatures, and this metric tracks the time for one replica to go from the lowest to the highest temperature and back again. A short round-trip time indicates that the simulation is sampling conformations efficiently, which is necessary to obtain reliable results. REMD round-trip time Lifetime curve data is a graph that shows how long a specific molecular interaction, such as a hydrogen bond, typically lasts in a simulation. The curve usually plots the probability that an interaction, once formed, will still exist after a certain amount of time has passed. This data is generated by calculating the autocorrelation function of an interaction's existence over a trajectory, and fitting the curve provides a characteristic lifetime for that interaction. lifetime curve data An average taken over a window that moves along a series, giving one value for each position of the window. It removes short-term fluctuation so that a slower trend can be seen. The width of the window sets what counts as fluctuation and what counts as trend. (UNVERIFIED) running average A measure of how far 2 sets of collective modes span the same motions, computed from the inner products between their eigenvectors. A value near 1 means the 2 sets describe the same motion, and a value near 0 means they do not. It is how the modes from 2 simulations, or from a simulation and an experiment, are compared. (UNVERIFIED) root-mean-square inner product A single number summarising how strongly the parts of a molecule move together, computed as the root mean square of all the pairwise correlation values. It reduces a whole correlation matrix to 1 figure, so 2 simulations can be compared directly. It says how much correlated motion there is but not where. (UNVERIFIED) RMS average correlation An interaction energy density giving, for each cell of a 3-dimensional grid around a solute, the mean interaction energy between the solute and the solvent molecules in that cell, per unit volume. Negative regions mark places where solvent is held favourably by the solute. Because it is per volume, it reflects both how strongly the solvent there interacts with the solute and how much solvent is present. (UNVERIFIED) solute-solvent interaction energy density An interaction energy density giving, for each cell of a 3-dimensional grid around a solute, the mean interaction energy of the solvent molecules in that cell with all other solvent molecules in the system, per unit volume. Comparing it with the bulk value shows where a solute has disturbed the solvent's own network of interactions. Solvent that interacts less well with other solvent than in the bulk is a sign of a place where a ligand could displace it favourably. (UNVERIFIED) solvent-solvent interaction energy density An energy giving the mean non-bonded energy of each solvent molecule found in a small region of space, such as 1 cell of a grid around a solute, with electrostatics computed by particle mesh Ewald. It includes the molecule's interactions with both the solute and the other solvent molecules, and it is obtained by dividing the PME solvent energy density of the region by the number density of solvent there. Comparing it with the bulk value shows where solvent is more or less stable than in pure solvent. (UNVERIFIED) normalized PME solvent energy An energy density giving, for each cell of a 3-dimensional grid around a solute, the mean non-bonded energy of the solvent molecules in that cell per unit volume, with electrostatics computed by particle mesh Ewald. It includes the solvent's interactions with both the solute and the rest of the solvent, so it shows where solvent is energetically stabilised and where it is strained. It differs from a solvent density map in reporting energy rather than how much solvent is present. (UNVERIFIED) PME solvent energy density An interaction energy giving the mean interaction energy of a water molecule with all the other water molecules, expressed per water molecule found in a given region, such as 1 cell of a grid around a solute. Expressing it per molecule makes regions holding different amounts of water comparable with each other and with bulk water. A value less negative than in bulk marks water that is held less strongly by the surrounding water than in the bulk. (UNVERIFIED) normalized water-water interaction energy An electronic property giving, at each point of a 3-dimensional grid around a solute, the mean dipole moment of the solvent molecules there per unit volume. It shows how solvent molecules orient themselves around a solute, which a plain density map cannot. Regions where the dipoles all point the same way mark strong local electric fields. (UNVERIFIED) dipole moment density A structural property giving, for each cell of a 3-dimensional grid around a solute, the mean number of neighbours of the solvent molecules in that cell multiplied by the number of solvent molecules per unit volume there. It therefore combines how crowded each molecule's surroundings are with how much solvent the cell holds. Dividing it by that number density gives the neighbour count per molecule. (UNVERIFIED) neighbor density A coordination number giving the average number of neighbouring solvent molecules, usually water, around each solvent molecule found in a small region of space, such as 1 cell of a grid around a solute. Dividing by the number of solvent molecules in that region makes it a count per molecule, so regions holding different amounts of solvent can be compared. In grid inhomogeneous solvation theory (GIST), 2 water molecules count as neighbours when their oxygen atoms are within 3.5 Å of each other. (UNVERIFIED) normalized neighbor count A measure of how closely the 4 nearest neighbours of a water molecule sit at the corners of a regular tetrahedron around it. Water in its ordinary hydrogen-bonded network is close to tetrahedral, and disruption shows as a fall in the value. It is how water near a surface or an ion is shown to be structured differently from bulk water. (UNVERIFIED) tetrahedral order parameter A tetrahedral order parameter rescaled to run from −1/3 to 1, where 1 is a perfect tetrahedron and 0 is a random arrangement. The rescaling puts the value of random water at 0, so a number can be read directly as how much more ordered than random the arrangement is. It is the form usually reported. (UNVERIFIED) normalized tetrahedral order parameter A chemical shift restraint simulation is a method where experimental Nuclear Magnetic Resonance (NMR) chemical shifts are used to guide a molecular dynamics simulation. An additional energy term is added that penalizes the simulation if the back-calculated chemical shifts from the trajectory do not agree with the experimental values. For experts, this method is a powerful tool for refining the accuracy of a structural ensemble to be consistent with experimental NMR data. chemical shift restraint simulation A pseudocontact shift restraint simulation is a technique that uses a specific type of NMR data, known as pseudocontact shifts (PCS), to guide a simulation. PCS provide long-range distance and angular information relative to a paramagnetic metal center in the molecule. For structural biologists, using PCS restraints is a powerful method for determining the structure and dynamics of proteins and their complexes. pseudocontact shift restraint simulation A direct dipolar coupling restraint simulation is a method that incorporates experimental NMR data from residual dipolar couplings (RDCs) to guide a MD simulation. RDCs provide information about the orientation of chemical bonds relative to an external magnetic field. direct dipolar coupling restraint simulation A residual chemical shift anisotropy (RCSA) restraint simulation is a technique that uses NMR RCSA data to guide a MD simulation. RCSA provides information about the orientation of the chemical shift tensor for specific atoms, which is sensitive to local structure and dynamics. residual chemical shift anisotropy restraint simulation An energy term printed in a simulation's energy output that tracks part of an algorithm rather than any physical interaction. Thermostats and barostats that add extra degrees of freedom report such terms, as does the correction term of adaptive QM/MM, so that a conserved total can still be checked. Such a term is not part of the potential energy and must not be added to it. (UNVERIFIED) book-keeping energy term An electrostatic method that represents the solvent beyond a cutoff distance as a uniform medium that polarizes in response to the charges inside, rather than as individual molecules. The polarized medium pushes back with a field that is added to the direct interactions. It is much cheaper than Ewald summation and less accurate, particularly for a system carrying a net charge. (UNVERIFIED) solvent reaction field potential A collective variable is a simplified coordinate that helps describe a complex process in a simulation, such as the distance between two proteins as they bind. Instead of tracking thousands of atomic coordinates, a collective variable reduces the system's motion to one or a few key parameters that capture the essence of the event. These variables are frequently used in enhanced sampling methods to focus computational effort on exploring a specific conformational change or reaction pathway. collective variable A distance collective variable is a simple coordinate that measures the separation between two specific atoms or groups of atoms. It is one of the most common variables used in simulations to track processes like a drug molecule unbinding from its target protein. Monitoring this distance provides a straightforward way to follow the progress of an association or dissociation event. distance collective variable A center of mass distance collective variable measures the separation between the centers of mass of two groups of atoms, such as two entire proteins or a protein and a ligand. This provides a robust measure of the distance between whole molecular entities, averaging out the fast motions of individual atoms. center of mass distance collective variable This is a complex coordinate that tracks the difference between two separate center-of-mass distance measurements. It defines a reaction coordinate by calculating the absolute difference between two distinct COM distance measurements. difference of center of mass distances collective variable A linear combination of distances is a mathematical expression used in molecular simulations to define a collective variable (CV). It sums or subtracts multiple specific atomic distances, each multiplied by a defined weight, creating a single reaction coordinate to track conformational changes like folding or chemical reactions. linear combination of distances collective variable An angle collective variable is a low-dimensional geometric descriptor that measures the angle formed by three specific atoms or groups of atoms during a simulation. This variable is used to monitor hinge-like motions, such as the opening and closing of domains in a protein. angle collective variable A center of mass angle collective variable measures the angle formed by the centers of mass of three distinct groups of atoms. This is useful for describing the large-scale orientation of entire protein domains or molecules relative to one another, smoothing out the rapid fluctuations of individual atoms. center of mass angle collective variable A torsion collective variable is a coordinate that tracks the value of a specific dihedral angle, which describes the rotation around a chemical bond. These variables are fundamental for describing the detailed three-dimensional shape of a flexible molecule. They are often used in enhanced sampling simulations to accelerate the rotation of protein side chains or to explore different backbone conformations. torsion collective variable A center of mass torsion collective variable is a dihedral angle defined by the centers of mass of four distinct groups of atoms, rather than four single atoms. This allows for the description of torsional or twisting motions between large molecular domains or subunits. center of mass torsion collective variable This is a collective variable created by summing the cosine values of several different dihedral angles within a molecule. This mathematical combination creates a smooth, continuous variable that describes the collective rotational state of multiple bonds at once. It is often used in enhanced sampling methods to create a simplified coordinate that captures a complex, multi-dimensional conformational change. sum of cosines of dihedrals collective variable This is a collective variable created by summing the sine values of several different dihedral angles within a molecule. This mathematical combination creates a smooth, continuous variable that describes the collective rotational state of multiple bonds at once. It is often used in enhanced sampling methods to create a simplified coordinate that captures a complex, multi-dimensional conformational change. sum of sines of dihedrals collective variable A paired dihedral collective variable is a coordinate that combines two different dihedral angles, often the phi and psi backbone angles of a protein, into a single analysis. This is most commonly used to generate a two-dimensional free energy map, known as a Ramachandran plot, which reveals the energetically allowed conformations of the protein backbone. This analysis is fundamental for understanding protein structure and flexibility. paired dihedral collective variable A pattern dihedral is a specialized Collective Variable (CV) used in MD to track the conformational transitions of a molecular system, such as protein folding or crystal polymorphs. It measures a weighted linear combination of multiple specific dihedral angles to act as a low-dimensional structural parameter, often replacing a complex set of atomic coordinates. pattern dihedral collective variable The radius of gyration is a collective variable that measures the overall compactness of a molecule or a group of atoms. It quantifies how spread out the atoms are from the molecule's center of mass, with a smaller value indicating a more compact shape. This variable is commonly used as a simple metric to monitor large-scale conformational changes, such as the process of a protein folding into its native structure or unfolding. radius of gyration collective variable A multi-RMSD collective variable is a coordinate that simultaneously measures a molecule's structural similarity to several different reference structures. This is useful for tracking a molecule's transitions between multiple known functional states, such as an active, inactive, and an intermediate conformation. By monitoring which reference structure the molecule is closest to, this variable can map out the complex conformational landscape connecting different states. multi-RMSD collective variable A number of bonds collective variable is a coordinate that counts the number of specific chemical bonds that exist at any given moment in a simulation. This variable is only meaningful when using a reactive force field that allows chemical bonds to form and break. It serves as a simple and direct way to track the progress of a chemical reaction by monitoring the creation or destruction of key covalent bonds. number of bonds collective variable A handedness collective variable is a mathematical measure that quantifies the chirality, or "handedness," of a molecular structure, such as the left-handed versus right-handed twist of a helix. This variable provides a continuous value that can track a molecule's transition between different chiral states during a simulation. handedness collective variable A complex quantity describing how strongly and with what phase a crystal scatters a wave in a given direction, summed over the atoms in its repeating unit. Its magnitude is what a diffraction experiment measures; its phase is not, which is the central difficulty of crystallography. The electron density follows from both by Fourier transform. (UNVERIFIED) structure factor A physical computing device or component, such as a processor, a whole computer or cluster, or the network between machines, on which simulation software runs. It is recorded with a simulation because it determines how long the run takes and can decide whether the run can be repeated bit for bit. Averaged scientific results should not depend on the hardware used. (UNVERIFIED) hardware A parameter giving the effective size of an atom, as the distance at which another atom begins to be repelled strongly. It is the radius of the hard sphere that best represents where Pauli repulsion overtakes the weak attraction between 2 non-bonded atoms. These radii set the molecular surface used in solvation models and in most measures of packing. (UNVERIFIED) van der Waals radius The Larmor frequency is the specific rate at which a magnetic particle, like an atomic nucleus, wobbles or "precesses" when placed in a strong external magnetic field. This is analogous to the way a spinning top wobbles in the Earth's gravitational field. This frequency is the fundamental principle behind all Nuclear Magnetic Resonance (NMR) experiments, as it determines the exact frequency of energy the nucleus will absorb. Larmor frequency A function that varies smoothly between 0 and 1 across a shell of space, used to blend an atom's quantum and classical descriptions as it crosses the boundary. Blending removes the jump in energy that an abrupt reassignment would cause. The width of the shell trades smoothness against how many atoms are in a mixed state at once. (UNVERIFIED) QM/MM switching function A simulation analysis that tests whether a simulation has run long enough for the quantity of interest to stop drifting. It follows a chosen property, such as total energy or a structural measure, and checks whether its running average settles over the later part of the trajectory. Without it, a reported average may reflect only the part of the trajectory that happened to be sampled. (UNVERIFIED) convergence analysis A modeling and setup tool is a type of software used to construct and prepare a molecular system before a simulation is run. This involves tasks like building a molecule from scratch, repairing a structure from an experimental database, or surrounding a protein with water molecules to mimic its natural environment. For an expert, these tools are essential for creating a chemically correct and physically stable starting point, which is a prerequisite for any meaningful simulation. modeling and setup tool A software program that performs the calculation at the centre of a simulation, computing the energy of a molecular system and the forces on its atoms and, in a dynamics run, advancing the atom positions step by step. It takes a prepared system and a set of parameters and produces energies, optimised structures or a trajectory. Other programs in a simulation workflow mostly prepare its input or analyse its output. (UNVERIFIED) simulation engine A dynamics analysis method that reduces the motion in a trajectory to the few collective movements that account for most of the total fluctuation, using principal component analysis. The largest components are usually functionally meaningful, such as hinge bending between domains, while the smallest are thermal noise. It lets a trajectory of thousands of coordinates be discussed in terms of 2 or 3 motions. (UNVERIFIED) essential dynamics analysis A matrix whose entry at row i and column j gives the correlation between the fluctuations of quantities i and j. In trajectory analysis the quantities are usually the displacements of atoms or residues, so the matrix says which parts of the molecule move together. It is the input from which correlated motion and network analyses are computed. (UNVERIFIED) correlation matrix A structural analysis method that identifies which parts of a molecule are rigid and which move, usually by computing the root-mean-square fluctuation of each atom or residue about its average position. The result separates a stable core from mobile loops and termini. Its results can be compared with experimental B-factors, which measure atomic mobility in crystal structures. (UNVERIFIED) flexibility analysis A contact map is a simplified 2D picture that shows which parts of a molecule are touching other parts of itself or other molecules. Contact maps provide a representation of molecular interactions or a protein foldings. contact map A contact map recording which pairs of atoms or residues are in contact in a reference folded structure. It is built once from an experimental structure and then used as the benchmark a simulated conformation is scored against. The fraction of its contacts that a conformation reproduces is the usual measure of how folded that conformation is. (UNVERIFIED) native contact map A contact map recording which pairs of atoms or residues are in contact in a given conformation but not in the reference folded structure. It shows interactions that form only in unfolded, misfolded or intermediate states during a simulation. Such contacts can hold a protein in a wrong shape and slow its folding. (UNVERIFIED) non native contact map Hydrophobic patches are clusters of non-polar, "water-fearing" atoms or amino acid side-chains that are exposed on the surface of a larger molecule, such as a protein. Because water molecules are highly polar, they are repelled by these non-polar regions and instead cluster together to interact with themselves. hydrophobic patches hydrophobic patch A simulation analysis that represents a molecule as a graph, with atoms or residues as nodes and interactions or correlations as edges, and then applies graph theory to it. It finds residues that act as hubs and traces paths of correlated motion between distant sites. It is widely used to study allosteric communication, the way a change at a site affects a distant one, and the long-range effects of mutations. (UNVERIFIED) network analysis A structural analysis method that finds cavities on or inside a molecule and measures their size, shape and how they change over time. A cavity that appears only briefly cannot be seen in a single experimental structure. Finding such transient cavities is one of the things simulation adds to a static structure. (UNVERIFIED) pocket analysis A simulation analysis that examines the behaviour of a small molecule bound to a larger one during a simulation. It covers how the ligand moves within its site, which contacts it makes and keeps, and whether it stays bound at all. (UNVERIFIED) ligand analysis A complete assembly of hardware on which a simulation runs, from a single workstation to a large cluster or a rented cloud platform. It is recorded because how long a simulation takes, and sometimes whether it can be repeated bit for bit, depends on it. Averaged scientific results should not depend on the machine used. (UNVERIFIED) computing system An analysis method that characterises the pocket a ligand occupies, measuring its volume, shape and how that changes as the simulation proceeds. A pocket that opens and closes during a run cannot be judged from the single structure a crystal gives. It is how a binding site is shown to be druggable or transient. (UNVERIFIED) ligand cavity analysis Residence time is the average amount of time that a molecule, such as a water molecule or a ligand, spends in a specific, defined location, like a protein's binding site. It is calculated from a simulation by monitoring when a molecule enters a region and when it leaves, and then averaging this duration over many such events. residence times residence time A confidence score that a structure prediction model gives each residue, on a scale from 0 to 100, estimating how well the distances between that residue and its near neighbours match the true structure. It is the model's own prediction of the local distance difference test (lDDT) score, so a confident prediction can still be wrong. Regions scoring low are often genuinely disordered rather than badly predicted. (UNVERIFIED) pLDDT Backbone DH fluctuations are a measure of how much the backbone of a protein wiggles and changes shape during a simulation. It specifically quantifies the fluctuations in the dihedral angles (DH) of the protein's backbone, which describe the rotations around the chemical bonds that connect the amino acids. Analyzing these fluctuations provides a detailed picture of the protein's flexibility and dynamics that is independent of its overall tumbling motion. backbone DH fluctuations backbone DH fluctuation Membrane analysis is a category of analysis methods specifically focused on characterizing the properties of lipid bilayer simulations. These methods are used to quantify the structural and dynamic behavior of the membrane itself and its interactions with other molecules, like embedded proteins. This includes calculating key properties such as the area per lipid, membrane thickness, and lipid order parameters to validate the simulation and understand membrane function. membrane analysis The area per lipid is a fundamental biophysical metric representing the average cross-sectional surface area occupied by a single lipid molecule within a lipid bilayer. It is calculated by dividing the total area of the simulation box in the membrane plane by the number of lipids in one leaflet. area per lipid Lipid contact analysis is an analysis method used to identify and quantify the interactions between lipid molecules and a protein embedded in the membrane. The analysis tracks which specific lipid molecules are in close contact with the protein over the course of a simulation. This method may be used for identifying potential "hotspot" lipid binding sites on a membrane protein and understanding the composition of the annular lipid shell. lipid contact analysis Lipid density is a measure of how tightly packed the lipid molecules are in a simulated cell membrane. It is typically calculated as a one-dimensional profile along the axis perpendicular to the membrane plane, showing the mass or electron density of different lipid components like headgroups and tails. Analyzing the lipid density profile is a fundamental check to ensure the simulated membrane has the correct structure and thickness, and it reveals how an embedded protein might locally perturb the lipid packing. lipid density Lipid distribution refers to the spatial arrangement and composition of different types of lipid molecules within a simulated membrane. In a mixed bilayer, lipids are often not randomly distributed but can form distinct domains or clusters, and some lipids may preferentially accumulate around an embedded protein. Analyzing this distribution is crucial for understanding membrane organization and the specific lipid environment required for a membrane protein's function. lipid distribution analysis Lipid order parameters are a set of values that measure how well-aligned and ordered the flexible tails of lipid molecules are within a simulated cell membrane. A value of 1 means the lipid tail is perfectly straight and aligned perpendicular to the membrane surface, while a value of 0 means it is completely disordered. These parameters, specifically the deuterium order parameter (SCD), can be directly compared with experimental results from NMR spectroscopy, making them one of the most important metrics for validating the physical accuracy of a membrane simulation. lipid order parameters lipid order parameter Membrane geometry describes the overall three-dimensional shape and dimensions of a simulated cell membrane. This includes properties like the membrane's thickness, its lateral dimensions in the simulation box, and its overall shape, which can be planar, curved, or even vesicular. Characterizing the membrane geometry is a fundamental part of analyzing membrane simulations, ensuring the system is stable and providing the context for the behavior of embedded proteins. membrane geometry Membrane curvature is a measure of how much a simulated cell membrane is bent or curved. This property is calculated by fitting the positions of the lipid headgroups to a mathematical surface and then determining the local or average curvature of that surface. Analyzing membrane curvature is essential for studying processes like vesicle formation, membrane fusion, and the influence of proteins that induce or sense bending in the lipid bilayer. membrane curvature The membrane width is the thickness of a simulated cell membrane, measured from one side to the other, along the z-axis. It is typically defined as the distance between the peaks in the density profile of the lipid headgroups or phosphate atoms on opposite sides of the bilayer. The membrane width is a key structural parameter that is monitored to confirm a simulation is properly equilibrated and is used to understand how membrane proteins match their hydrophobic length to the surrounding lipid environment. membrane width The pore dimension refers to the size and shape of a channel or tunnel that passes through a protein. This property is calculated by finding the radius of the narrowest point (the constriction site) and the overall profile of the channel's radius along its length. Analyzing the pore dimensions is critical for understanding how ion channels and transporters function, as it determines which molecules are small enough to pass through the membrane. pore dimension DNA bendability is the intrinsic mechanical property that describes how easily a DNA double helix can deviate from its ideal straight conformation by adopting a bent or curved structure. This property depends on the nucleotide sequence, with some sequences being inherently more flexible than others. Sequence-dependent bendability plays a key role in DNA packaging, nucleosome formation, and protein-DNA recognition, as many proteins induce or stabilize DNA bending upon binding. DNA-bendability Base breathing refers to a specific and transient opening of one or more base pairs in a DNA or RNA double helix due to the temporary disruption of hydrogen bonds. During these events, the bases become transiently exposed to the surrounding solvent and more accessible without necessarily becoming fully unstacked from the helix. When this opening occurs at the ends of the duplex, it is commonly referred to as terminal fraying. These spontaneous opening events are a natural part of the helix's dynamics and are thought to be important for allowing proteins to access the genetic information encoded in the bases. Analyzing the frequency and duration of these motions provides insight into the local stability and accessibility of the DNA or RNA structure. base breathing A helical parameter is one of a set of measurements that describe the detailed 3D shape of a DNA or RNA double helix. These parameters quantify the relative position and orientation of adjacent base pairs, including values for twist, roll, slide, and rise, which define the local helical geometry. Calculating these parameters from a simulation trajectory is the standard method for characterizing the sequence-dependent structure of nucleic acids and for distinguishing between different helical forms like A-DNA and B-DNA. helical parameter Inter-base pair (Inter-BP) correlation analysis refers to the statistical coupling between the structural parameters of neighboring DNA base pairs or base-pair steps. It quantifies how variations in one helical parameter (e.g., roll, twist, or slide) are associated with changes in the same or different parameters at adjacent base pairs. These correlations reveal sequence-dependent structural coupling and cooperative conformational behavior within the DNA double helix, providing insight into its flexibility, mechanics, and recognition by DNA-bining proteins. inter-BP correlation analysis Nucleic acid backbone flexibility is a measure of the ability of the sugar-phosphate backbone of DNA or RNA to adopt different conformations through rotations about its backbone torsion angles. It is commonly characterized by analyzing the distributions and fluctuations during a simulation of the backbone dihedral angles (α, β, γ, δ, ε, ζ), the glycosidic angle (χ), and the sugar pucker. Backbone flexibility plays a fundamental role in determining the structural dynamics, mechanical properties, and biological functions of nucleic acids, influencing DNA bending and packaging as well as RNA folding and molecular recognition. nucleic acid backbone flexibility Nucleosome wrapping describes the way a long DNA molecule is spooled around a histone protein octamer, forming the fundamental repeating unit of chromatin. Studying nucleosome wrapping is essential for understanding how DNA is compacted and how access to the genetic code is regulated for gene expression. nucleosome wrapping A property giving how strongly a structure resists being deformed by an applied force. For a single bond or angle it is set by the force constant in the force field; for a whole molecule or assembly it emerges from the network of interactions. It is the inverse of flexibility and is measured in force per unit displacement. (UNVERIFIED) stiffness A dynamics analysis method that identifies groups of atoms whose movements are correlated, so that they travel together as a unit rather than fluctuating independently. It picks out large-scale events such as a protein domain shifting or twisting as a whole. Those events are usually the ones that matter for function, such as an active site opening and closing. (UNVERIFIED) concerted motions analysis Trajectory meta-analysis refers to a higher level of analysis that involves comparing, contrasting, or combining the results from multiple, independent simulation trajectories. This can be used to improve statistical certainty, assess the convergence of a simulation, or identify the differences in dynamics between a wild-type protein and a mutant. This approach is essential for drawing robust conclusions that are not dependent on a single simulation run. trajectory meta analysis Trajectory comparison is a specific type of meta-analysis where two or more simulation trajectories are quantitatively compared to identify similarities and differences in their conformational sampling. This is often done by calculating metrics that measure the overlap of the conformational ensembles or by projecting the trajectories onto a common free energy landscape. This method is used to rigorously assess the impact of a mutation or ligand binding on a protein's dynamics. trajectory comparison A molecular simulation system is the complete collection of all the molecules included in a computational model for a simulation. This includes the main molecule of interest, such as a protein, as well as all the surrounding solvent molecules, ions, and any other components. molecular simulation system A carbohydrate system is a type of molecular simulation system where the primary molecules being studied are carbohydrates. This can range from a single sugar molecule in water to a complex polysaccharide or the glycan portion of a glycoprotein. These systems require specialized force fields and analysis techniques to accurately model the unique stereochemistry and flexibility of glycosidic linkages. carbohydrate system A complex system in this context refers to a molecular simulation system that contains two or more interacting macromolecules, or a macromolecule interacting with a smaller molecule like a drug. The primary focus of simulating such a system is to study the interactions, binding, and functional consequences of the molecules coming together. complex system A ligand complex system is a type of complex system where the focus is on the interaction between a large macromolecule, such as a protein or DNA, and a smaller molecule, known as a ligand. These simulations are central to the field of drug discovery, where the goal is to understand how a potential drug molecule binds to its biological target. ligand complex system A DNA-ligand system is a type of molecular simulation system where the primary molecules being studied are DNA molecules interacting with small molecule ligands. These simulations are used to study how drugs or other small molecules bind to DNA, which can be a mechanism for anti-cancer or anti-viral therapies. These systems are important for understanding intercalation, groove binding, and the structural changes induced in DNA upon ligand binding. DNA-ligand system An RNA-ligand system is a type of molecular simulation system where the primary molecules being studied are RNA molecules interacting with small molecule ligands. With the growing recognition of RNA as a major drug target, these simulations are becoming increasingly important for discovering and designing new therapeutics. These systems are challenging due to the high flexibility of RNA and the often subtle nature of the binding interactions. RNA-ligand system A molecular system consisting of a protein bound to a small molecule. Such systems are central to computer-aided drug design, because most drugs act by binding to a protein. What is usually studied is how tightly the 2 bind and which contacts hold them together. (UNVERIFIED) protein-ligand system A hybrid complex system is a type of complex system composed of macromolecules from different biological classes. This most commonly refers to a complex formed between a protein and a nucleic acid, such as a transcription factor bound to DNA. Simulating these systems is essential for understanding many fundamental biological processes like gene regulation and DNA repair. hybrid complex system A nucleic acid complex system is a type of molecular simulation system where the primary molecules being studied are two or more nucleic acid molecules. This can include a standard DNA double helix, an RNA duplex, or more complex structures like DNA-RNA hybrids or multi-stranded RNA junctions. These simulations are crucial for understanding the structure and stability of the genetic material and functional RNA molecules. nucleic acid complex system A DNA-DNA system is a type of molecular simulation system where the primary molecules being studied are two or more interacting strands of DNA. The most common example is the simulation of a DNA double helix to study its sequence-dependent structure and dynamics. These systems are also used to study more complex assemblies like DNA quadruplexes or Holliday junctions. DNA-DNA system A DNA-RNA system is a type of molecular simulation system where the primary molecules being studied are one ore more DNA strands interacting with one or more RNA strands. Such hybrid helices are important biological intermediates that occur during processes like transcription and reverse transcription. Simulating these systems is important for understanding the structural basis of these fundamental genetic processes. DNA-RNA system An RNA-RNA system is a type of molecular simulation system where the primary molecules being studied are two or more interacting RNA strands. This includes RNA duplexes, which are common in viral genomes and regulatory RNAs, as well as the vast and complex RNA assemblies that form the ribosome. Simulating these systems is critical for understanding the structural basis of RNA function and catalysis. RNA-RNA system A molecular system in which a protein is bound to another macromolecule, either a second protein or a nucleic acid. Such a system is simulated to study how the 2 recognise each other and what holds them together. The interface between them is usually what the study is about. (UNVERIFIED) protein complex system A protein-DNA system is a type of molecular simulation system where the primary molecules being studied are proteins interacting with DNA molecules. These systems are fundamental to nearly all aspects of genetics, including DNA replication, repair, and the regulation of gene expression by transcription factors. Simulating these complexes is crucial for understanding the molecular basis of how proteins read and manipulate the genetic code. protein-DNA system A protein-RNA system is a type of molecular simulation system where the primary molecules being studied are proteins interacting with RNA molecules. These complexes, known as ribonucleoproteins, are central to a vast range of cellular processes, including protein synthesis by the ribosome and RNA splicing. Simulating these systems is essential for understanding the intricate interplay between proteins and RNA that governs much of cell biology. protein-RNA system A protein complex system made up of 2 or more proteins bound to each other. Such pairings carry most cellular signalling and hold most large cellular structures together. What is usually studied is the interface: which residues touch, and how tightly. (UNVERIFIED) protein-protein system A molecular system whose subject is the behaviour of ions, usually dissolved in water. Such systems are simulated to test how well a force field reproduces measured hydration free energies, activity coefficients and ion pairing. Getting ions right is a precondition for trusting any simulation of a charged biomolecule. (UNVERIFIED) ion system A lipid system is a type of molecular simulation system where the primary molecule being studied are lipids, usually arranged as a bilayer in water. These simulations are used to study the fundamental physical properties of biological membranes, such as their phase behavior, fluidity, and permeability. lipid system A nucleic acid system is a type of molecular simulation system where the primary molecules being studied are nucleic acids, such as a DNA duplex or a folded RNA. These simulations are typically performed to study the intrinsic structure, flexibility, and conformational dynamics of the nucleic acid. nucleic acid system A DNA system is a type of molecular simulation system where the primary molecules being studied are DNA. This most often involves the simulation of a DNA double helix to investigate its sequence-dependent structure, flexibility, and dynamics. These simulations are used to understand properties like DNA bending and the transitions between different helical forms like A- and B-DNA. DNA system An RNA system is a type of molecular simulation system where the primary molecules being studied are RNA. Unlike DNA, RNA molecules often fold into complex and specific three-dimensional structures to perform their biological functions. Simulating these systems is crucial for understanding the principles of RNA folding and the structural basis of RNA catalysis. RNA system A protein system is a type of molecular simulation system where the primary molecules being studied are proteins. These simulations are typically run to study its structural stability, intrinsic flexibility, and conformational changes. This is the most common type of simulation, fundamental to understanding how a protein's structure relates to its function. protein system An Intrinsically Disordered Protein (IDP) system refers to proteins or regions of proteins that lack a fixed, stable three-dimensional structure under physiological conditions. Instead of adopting a single rigid shape, they exist as dynamic ensembles of interconverting structures that drive vital cellular functions. IDP system intrinsically disordered protein system A membrane protein system is a specific type of protein system where the protein is embedded within a lipid bilayer, mimicking its native environment in a cell membrane. These are among the most complex and computationally demanding types of simulations to set up and run. For experts, these simulations are essential for understanding the function of proteins that act as channels, transporters, and receptors in the cell. membrane protein system A molecular system whose subject is one or more small organic molecules rather than a macromolecule. Such systems are simulated to obtain properties like solubility or partitioning, and to test force field parameters before they are used on anything larger. They are small enough that long simulations and many repeats are affordable. (UNVERIFIED) small molecule system A molecular system containing water and nothing else. It is simulated to test how well a water model reproduces measured properties such as density, diffusion coefficient, dielectric constant and the temperature of maximum density. No water model reproduces all of them at once, so the choice depends on which matter for the intended study. (UNVERIFIED) water system A property of a simulation model that gives how much detail is kept when the real system is represented, from every atom as a separate site down to whole molecules or domains as single sites. It fixes both what the model can show and how much a simulation costs. Common levels are all-atom, united-atom, coarse-grained and mesoscopic. (UNVERIFIED) simulation model resolution All-atom is a simulation model resolution where every atom in the system, including all hydrogen atoms, is represented as an individual particle with its own parameters. This provides the highest level of structural detail and is the standard for most high-resolution biomolecular simulations. all atom A simulation model resolution in which several atoms are represented by 1 interaction site. Cutting the particle count and smoothing the energy landscape allows both larger systems and a longer timestep, so far longer times become reachable. Nothing smaller than a site can be observed in such a model. (UNVERIFIED) coarse grained A simulation model resolution in which whole proteins, or long segments of a polymer, are each represented by a single site. It is the coarsest level at which particles are still tracked individually, above which a continuum description is used instead. It is used for assembly and phase behaviour, where molecular detail does not affect the result. (UNVERIFIED) mesoscopic A simulation model resolution in which non-polar hydrogens are merged into the heavy atom they are bonded to, while all other atoms are kept. It reduces the number of particles, and so the cost, compared with an all-atom model, while keeping the hydrogens that take part in hydrogen bonds. It sits between all-atom and coarse-grained resolution. (UNVERIFIED) united-atom MDAnalysis is an open-source Python library used for the analysis of molecular dynamics simulation trajectories. It provides a flexible and object-oriented framework that allows users to read, write, and manipulate atomic coordinate data from a wide variety of simulation packages. MDAnalysis is a powerful tool for writing custom, complex, and reproducible analysis scripts in a familiar Python environment. MDAnalysis MDTraj is a modern, open-source Python library designed for the fast and efficient analysis of molecular dynamics trajectories. It excels at reading and writing a very broad range of trajectory and topology file formats, making it highly interoperable. MDTraj provides a fast and powerful toolkit for performing common structural and time-series analyses on simulation data. MDTraj A Python library for analysing protein structural dynamics, covering structure file parsing, conformational ensemble analysis and the calculation of collective modes. It brings normal mode analysis and elastic network models into a scripting environment. It is used to compare simulated motion with motion inferred from experimental structures. (UNVERIFIED) ProDy BioBB, or BioExcel Building Blocks, is a library of Python wrappers that create a standardized interface for many common biomolecular simulation programs like GROMACS and AmberTools. Each building block performs a specific task, such as running an energy minimization or calculating an RMSD, using a consistent input and output format. For computational biologists, BioBB provides an interoperable toolkit for constructing complex, reproducible, and portable simulation workflows. BioBB 3Dmol.js is a JavaScript library used for creating interactive, 3D visualizations of molecular structures directly within a web browser. It allows web developers to embed high-quality molecular viewers into websites and web applications without requiring any plugins. For creators of online scientific resources, 3Dmol.js provides a lightweight and versatile tool for displaying and interacting with molecular data on the web. 3DMol.js NGL Viewer is a WebGL-based JavaScript library for the fast and high-quality visualization of large molecular complexes and trajectories in the web browser. It is designed for high-performance rendering, enabling smooth, interactive exploration of even very large structures like ribosomes or viral capsids. NGL Open Babel is a versatile software program and library, often called a "chemical toolbox," for reading, writing, and converting between over 100 different chemical file formats. It allows users to easily interconvert between the various file types used by different simulation, modeling, and visualization programs. Its powerful interoperability makes it an essential utility in many simulation and modeling workflows for preparing and manipulating molecular structure files. Open Babel A system preparation program in the GROMACS suite that manipulates a structure file: converting between formats, setting the size and shape of the box, centring the molecule and rotating or translating it. It changes how a system is placed rather than what it contains. It is normally run before the box is filled with solvent. (UNVERIFIED) editconf A command-line program that provides access to every tool in the GROMACS suite through a single entry point, with the tool named as its first argument. It is not itself a simulation engine; it dispatches to the preparation, simulation and analysis programs. Almost every GROMACS command in a published protocol begins with it. (UNVERIFIED) gmx CHARMM-GUI is a powerful and widely used web server that provides a graphical interface for generating input files for a wide variety of complex biomolecular simulations. It automates many difficult and error-prone setup tasks, such as building lipid membranes, solvating proteins, and preparing systems for free energy calculations. CHARMM-GUI Modeller is a widely used computer program for generating three-dimensional models of proteins through homology or comparative modeling. It predicts a protein's structure by using the known structure of a related protein as a template and satisfying spatial restraints. For researchers, Modeller is a powerful tool to create high-quality structural hypotheses for proteins whose experimental structures have not yet been determined, providing a starting point for further simulation. Modeller A molecular dynamics engine, developed by Acellera, built to run biomolecular simulations on graphics processors, mainly NVIDIA ones. Its first version was presented as a way to reach microsecond timescales on a single workstation. Current versions are built on the OpenMM library. (UNVERIFIED) ACEMD A software platform for building, viewing and simulating molecular systems at the nanometre scale. Its functions are supplied by separate modules, so users and developers can add capabilities without changing the platform itself. That modular design is what distinguishes it from a fixed-feature suite. (UNVERIFIED) SAMSON Blender is a free and open-source 3D computer graphics software suite used for creating animated films, visual effects, art, and 3D printed models. While not a dedicated molecular viewer, it is often used by computational biologists to import molecular structures and create high-quality, cinematic renderings and scientific animations. Blender provides a professional-level tool for producing visually stunning figures and movies for publications and presentations. Blender Jmol is a free, open-source molecular viewer for 3D chemical structures, written in the Java programming language. Because it is Java-based, it can run on a wide variety of operating systems and was often used as a tool for displaying molecules on websites. For researchers and educators, Jmol provided a versatile and accessible tool for basic molecular visualization and analysis. Nowadays replaced by JavaScript and openGL tools such as JSMol, NGL or Mol*. Jmol MolStar (Mol*) is a modern, web-based toolkit for visualizing and analyzing large-scale molecular data. It is designed for high-performance rendering of extremely large structures, such as those from cryo-electron microscopy, directly in a web browser. For the structural biology community, Mol* is the state-of-the-art viewer integrated into databases like the PDB for exploring complex biological assemblies. Mol* MolStar A visualization and analysis program for molecular structures and related data, extensible through user-written modules. It produces publication-quality images and handles structures together with density maps and sequence data. It is no longer developed and has been replaced by its successor, UCSF ChimeraX, but it remains in use. (UNVERIFIED) UCSF Chimera A molecular modelling and simulation software suite covering structure prediction, structure cleaning, simulation and analysis in one program. Its steps are heavily automated, so a full workflow can be run without assembling separate tools. That automation is what distinguishes it from suites that expose each step separately. (UNVERIFIED) YASARA A computing system consisting of a group of individual computers (nodes) linked together by a high-speed network to function as a single, coordinated system for parallel processing. This encompasses a wide range of scales, from small laboratory clusters to massive institutional systems. (UNVERIFIED) HPC system high-performance computing cluster system A high-performance computing cluster system that operates at or near the highest performance available from any computer at the time it is installed. Such systems are very large, use specialised high-speed interconnects and usually serve as national or international resources for the most demanding computations. Examples include Frontier, and the Anton machines built only for molecular dynamics. (UNVERIFIED) supercomputer system A computing system made of virtual machines rented on demand over the internet from a provider's data centres, rather than owned by the user's institution. It lets researchers run large simulations, or use recent hardware such as high-end graphics processors, without buying and maintaining a machine of their own. (UNVERIFIED) cloud computing platform system The hardware component that performs the arithmetic and logic of a computation. A general-purpose one handles varied work a few operations at a time, while a graphics processor performs thousands of similar operations at once. Which kind a simulation runs on changes its speed by a large factor. (UNVERIFIED) processing unit The general-purpose processor of a computer, which handles varied instructions a few at a time with a large amount of fast local memory. It runs the parts of a simulation that are not easily done many at once, such as file handling and control logic. It is much slower than a graphics processor at the force calculation and much faster at everything irregular. (UNVERIFIED) central processing unit A processor holding thousands of small cores that carry out the same operation on many pieces of data at once. Computing the force on every atom is exactly that kind of work, which is why simulations run several times faster on one. It performs badly on work that branches differently for each item. (UNVERIFIED) graphics processing unit A processing unit designed for a single kind of computation, with its circuitry fixed when the chip is made. Specialising in this way gives a very large gain in speed and energy efficiency for that task, at the cost of flexibility. The Anton machines, built only for molecular dynamics, use such chips. (UNVERIFIED) ASIC application-specific integrated circuit Hardware that links the computers (nodes) of a cluster to one another so that they can exchange data at high speed. A simulation spread across many nodes exchanges data at every step, so the delay in that link can matter more than the speed of the processors. It is a main limit on how many nodes a single simulation can use efficiently. (UNVERIFIED) network interconnect A thermodynamic property giving the force per unit area that the particles exert on the boundaries of the simulation box. It is computed from the kinetic energy of the particles together with the virial, which combines the positions of the atoms with the forces between them. Its instantaneous value fluctuates much more than temperature, often by hundreds of atmospheres in a typical system, so only a long average is meaningful. (UNVERIFIED) pressure A thermodynamic property giving the size of the periodic simulation box. It is fixed in a constant-volume simulation and varies in a constant-pressure one, where it is the quantity the barostat adjusts. The mass in the box divided by the average volume gives the density that is compared with experiment. (UNVERIFIED) volume A networking technology used to connect computers in most local networks, and in faster versions also in many computing clusters. It is cheap and universal, but ordinary Ethernet delivers messages with more delay than a specialised cluster network. A simulation spread across many machines connected by ordinary Ethernet can spend much of its time waiting for data. (UNVERIFIED) ethernet A networking technology built for computing clusters, with very low delay and the ability to move data directly between the memory of different machines without involving their processors. That directness makes it suitable for a simulation whose parts must exchange data at every step. It is widely used as the interconnect of high-performance computing clusters. (UNVERIFIED) infiniband A networking technology developed by Intel for connecting the nodes of a supercomputer, offering low delay and high bandwidth in competition with InfiniBand. It was used in several large machines, and Intel has since stopped developing it. (UNVERIFIED) omni-path MDRepo is a public repository dedicated to the archival and sharing of molecular dynamics simulation data generated by the research community. It provides a standardized infrastructure for scientists to upload, store, and retrieve simulation trajectories and associated metadata. By centralizing access to these datasets, MDRepo promotes the reproducibility and secondary analysis of biomolecular simulations. MDrepo ATLAS is a specialized database that stores a comprehensive collection of standardized all-atom molecular dynamics simulations of proteins. It combines simulation trajectories with analyses of protein flexibility and dynamics, linked to experimental structural data. This resource allows researchers to explore the dynamic properties of a wide range of protein families in a consistent and comparative manner. ATLAS GPCRmd is a community-driven database and web platform specifically focused on molecular dynamics simulations of G-protein-coupled receptors (GPCRs). It provides a repository for storing simulation trajectories of this important protein family, along with tools for interactive visualization and analysis. This resource aids researchers in understanding the complex signaling mechanisms and conformational landscapes of GPCRs. GPCRmd InflateGRO is a utility script used to construct lipid membrane systems by packing lipids around a membrane protein. It works by initially expanding the distance between lipid molecules and then iteratively shrinking the system back to a realistic density while removing lipids that overlap with the protein. This method automates the challenging process of embedding a protein structure into a tight, pre-equilibrated lipid bilayer. InflateGro MemProtMD is a database that provides a collection of coarse-grained molecular dynamics simulations for membrane proteins found in the Protein Data Bank. It uses an automated pipeline to insert protein structures into lipid bilayers and simulate their equilibrium dynamics. This resource offers insights into the membrane orientation, lipid interactions, and stability of a vast number of membrane proteins. MemProtMD A crystallographic build tool is a software utility used to reconstruct a full crystal lattice or supercell from the asymmetric unit provided in a structure file. It applies the specific symmetry operations defined by the space group to generate the complete crystallographic packing environment. crystallographic build tool A software program that defines the shape and size of a simulation box and fills it with solvent molecules and ions at a target density and concentration. It places the molecules so that none overlap, which a random arrangement would not achieve. It is the step that turns a lone molecule into a system with an environment. (UNVERIFIED) system packing and solvation tool A software program that standardises and repairs the formatting of a molecular structure file, by renumbering residues, correcting atom names, removing unwanted records and splitting or joining chains. The work is clerical rather than scientific, but nothing downstream runs until it is done. Most structure files from experimental sources need it. (UNVERIFIED) structure cleaning and manipulation tool A homology modeling tool is a software program designed to predict the three-dimensional structure of a protein based on its amino acid sequence and experimentally determined template structures. It constructs a model by mapping the target sequence onto the backbone of a homologous protein and computationally refining the loops and side chains to resolve steric conflicts. For experts, these tools are a solution for generating starting structural models when high-resolution crystallography or NMR data for the specific target is unavailable. homology modeling tool AutoDock Vina is an open-source software tool for molecular docking that predicts the preferred orientation and binding affinity of a small molecule to a protein receptor. It utilizes a sophisticated iterated local search global optimizer combined with an empirical scoring function to efficiently explore the conformational space of the ligand. AutoDock Vina A nucleic acid analysis tool is a specialized software program designed to characterize the geometry and topology of DNA and RNA structures. It calculates parameters such as base-pair step geometry (roll, tilt, twist), groove dimensions, and sugar pucker to describe the helical conformation. nucleic acid analysis tool A software program that computes the scattering pattern a structure or trajectory would produce in an experiment. Comparing that computed pattern with a measured one is how a simulated ensemble is tested against solution data. It turns coordinates into something directly comparable with a measurement. (UNVERIFIED) scattering analysis tool A software program that works out which acidic and basic groups in a molecule carry a proton, and at what acidity they gain or lose one. It either predicts those values from the structure or measures them from a simulation that allowed protons to move. Which groups are charged changes a molecule's behaviour, so these values matter both when a simulation is set up and when its results are read. (UNVERIFIED) titration analysis tool A software program that characterises the structure and thermodynamics of the solvent around a solute. It computes quantities such as radial distribution functions, solvent density maps, and the solvation free energies given by integral equation theories. It is how a hydration shell is described quantitatively rather than by eye. (UNVERIFIED) solvation analysis tool An algorithm and program that superposes 2 protein structures and scores how similar their folds are, on a scale that does not depend on how long the proteins are. Length independence is what lets a short and a long protein be compared meaningfully. It finds the correspondence between residues itself, so no alignment need be supplied. (UNVERIFIED) TM-align DALI (Distance Alignment) is a software method that compares protein structures by analyzing their internal distance matrices rather than their rigid-body superposition. This approach allows to detect distant evolutionary relationships and structural similarities even when the proteins have undergone significant conformational changes or distortions. DALI LS-align is an algorithm designed for atom-level structural comparison of ligand molecules. The target function of LS-align is a combination of inter-atom distance, atom mass, and chemical bond connections; while the final atom-to-atom alignment is generated by maximizing such target function through an enhanced-greedy based, iterative heuristic search algorithm. LS-align An enhanced sampling algorithm that manages multiple parallel simulation replicas running at different thermodynamic states (e.g., temperatures or Hamiltonians). It evaluates and periodically attempts to swap the states of these replicas based on a Metropolis-style acceptance criterion to facilitate barrier crossing and improve conformational sampling. replica exchange algorithm A software program that computes a free energy difference from simulations run along a path of non-physical intermediate states. It implements estimators such as the multi-state Bennett acceptance ratio or thermodynamic integration. It is run after the simulations, on the energies they recorded. (UNVERIFIED) alchemical analysis tool A free energy surface analysis tool is a software utility used to reconstruct the multidimensional free energy landscape of a system from enhanced sampling simulation data. It processes the history of visited states or applied biases, such as those from metadynamics or umbrella sampling, to generate a potential of mean force (PMF) map. Enable the visualization of low-energy basins and transition barriers along specific collective variables or reaction coordinates. free energy surface analysis tool A software program that inspects a molecule, works out which atom type each atom belongs to, and looks up the parameters that follow. The assignment is made from element and bonding environment. It is where a general force field becomes a specific set of numbers for one molecule. (UNVERIFIED) parameter assignment tool A software program that computes partial atomic charges for a molecule so it can be simulated with a force field. The method used must match the one the force field's own charges were derived with, or the molecule will interact too strongly or too weakly with everything else. It is run once per new molecule. (UNVERIFIED) charge derivation tool A software program that produces the library file defining a new building block: its atoms, their types and charges, the bonds between them and the atoms that join it to its neighbours. Once written, that template lets the block be used anywhere in a chain. It is what makes a non-standard residue reusable rather than a one-off. (UNVERIFIED) residue template tool A protein force field from the AMBER family, introduced by García and Sanbonmatsu in 2002, that removes some of the backbone phi and psi torsion terms of ff94 to reduce its overly strong preference for helical structure. It was developed to study the balance between helix and coil in peptides. (UNVERIFIED) AMBERGS A GROMOS parameter set whose partial charges were refitted so that amino acid side chain analogues reproduce measured hydration free enthalpies, while keeping the fit to solvation free enthalpies in cyclohexane from the 53A5 set. Fitting to both water and cyclohexane gives a more realistic balance between polar and non-polar environments. It is intended for simulations of biomolecules in water and became a common choice for protein simulation in that family. (UNVERIFIED) GROMOS 53a6 A GROMOS parameter set fitted to reproduce properties of biomolecules in non-polar surroundings, such as the interior of a membrane. It shares its bonded parameters with the aqueous-fitted set of the same generation and differs in the non-bonded ones. The 2 are alternatives chosen by environment, not successive versions. (UNVERIFIED) GROMOS 53a5 A GROMOS parameter set that extends GROMOS 43a1 with parameters for phosphorylated amino acids. Standard parameter sets do not include them, because phosphorylation is a modification rather than a standard residue. It covers phosphoserine, phosphothreonine and phosphotyrosine, each with the phosphate group doubly or singly charged. (UNVERIFIED) GROMOS 43a1p A force field of the GROMOS family, released in the late 1980s, in which non-polar hydrogens are merged into the heavy atoms they are bonded to. It set the parameter conventions the later GROMOS revisions were built on. It is of historical interest rather than in current use. (UNVERIFIED) GROMOS-87 PLUMED is an open-source software library designed to analyze molecular dynamics simulations and perform enhanced sampling of free energy landscapes. It functions as middleware that interfaces with various simulation engines to calculate collective variables and apply biasing potentials in real-time. PLUMED A parallel computing method that divides the simulation box into regions and gives each to a different processor, which handles the atoms currently inside it. Because interactions are short-ranged, a processor mostly needs data from its own region and its immediate neighbours. It is what allows a single simulation to use hundreds of processors at once. (UNVERIFIED) domain decomposition algorithm A family of united-atom force fields for biomolecules, developed with the GROMOS simulation software, in which non-polar hydrogens are merged into the heavy atoms they are bonded to. Its later parameter sets, such as 53A6 and 54A7, were fitted to reproduce measured liquid properties and free enthalpies of solvation in water and in cyclohexane. Each parameter set is named by a code such as 43A1 or 54A7. (UNVERIFIED) GROMOS force field The GROMOS96 force field is a family of united-atom molecular mechanics parameter sets developed in conjunction with the GROMOS96 simulation software. It treats non-polar hydrogen atoms as part of the heavy atoms they are bonded to (e.g., CH, CH2, CH3) to reduce computational complexity while maintaining accuracy for biomolecular simulations. GROMOS96 A coarse-grained force field that maps roughly 4 heavy atoms and their hydrogens onto 1 interaction bead, with beads classified by polarity, charge and hydrophobicity. The small number of bead types makes it transferable across lipids, proteins and carbohydrates. It is the most widely used coarse-grained model for membrane simulation. (UNVERIFIED) MARTINI force field A force field of the OPLS all-atom family with its protein backbone and side-chain torsion terms refitted again, by Robertson, Tirado-Rives and Jorgensen in 2015, to high-accuracy quantum energy scans of dipeptides. The new parameters were tested against measured properties of peptides in solution and improved on the earlier OPLS-AA/L revision. It is the latest OPLS version for proteins. (UNVERIFIED) OPLS-AA/M A force field of the OPLS all-atom family with its protein backbone and side-chain torsion terms refitted against high-level quantum calculations on peptides, published by Kaminski and co-workers in 2001. The refit improved the agreement of peptide conformational energies with the quantum results. It was the standard OPLS choice for proteins before the OPLS-AA/M revision. (UNVERIFIED) OPLS-AA/L A computer model of water that allows the electric charge distribution within each molecule to shift in response to its surroundings. Unlike fixed-charge models, it explicitly includes terms for electronic polarization, often via induced dipoles, Drude oscillators, or fluctuating charges. For experts, these models (e.g., TIP4P-FQ, AMOEBA water) capture non-additive many-body effects and are essential for accurately simulating interfaces, ion solvation, and spectroscopic properties. (UNVERIFIED) polarizable water An algorithm for systems with periodic boundary conditions in which, for short-range interactions, each particle interacts only with the nearest copy of every other particle, whether that copy is in the simulation box itself or in a neighboring periodic image. It keeps the number of interacting pairs finite in a system that repeats without end. It requires the cutoff radius to be less than half the shortest box vector, so that no particle interacts with 2 copies of the same particle or with its own image. (UNVERIFIED) minimum image convention A measurement of the closest separation between two distinct groups of atoms in a structure. It is calculated by evaluating the Euclidean distance between all possible atom pairs from the two groups and selecting the smallest value. This metric is frequently used in trajectory analysis to define contact maps, monitor collision avoidance, or serve as a collective variable for pathway sampling. minimum distance A free energy defined as the internal energy minus the product of temperature and entropy, A = U − TS, whose decrease equals the largest amount of work a system can do at constant temperature. At constant temperature and volume a process goes forward on its own only if it lowers this quantity. It is the free energy estimated by simulations at constant volume and temperature (NVT, the canonical ensemble), usually as a difference between 2 states by methods such as thermodynamic integration or the Bennett acceptance ratio. (UNVERIFIED) helmholtz free energy A free energy defined as the enthalpy minus the product of temperature and entropy, G = H − TS, whose decrease equals the largest amount of work other than expansion work that a system can do at constant temperature and pressure. At constant temperature and pressure a process goes forward on its own only if it lowers this quantity, so it decides the direction of change and the position of chemical equilibrium. It is the free energy estimated by simulations at constant pressure and temperature (NPT), for example for binding, solvation and phase equilibria. (UNVERIFIED) gibbs free energy A thermodynamic property computed as the sum, over all atoms, of each atom's position multiplied by the force acting on it, and used together with the kinetic energy to compute the pressure. In a periodic system it is computed from the distances and forces between pairs of atoms, and it is often reported as a 3-by-3 tensor from which the pressure tensor follows. Barostats rely on it, through the pressure, to decide how to change the box. (UNVERIFIED) virial A correlation function that measures how much the value of a quantity at one time resembles its own value a given time later, computed by averaging the product of the signal with a time-shifted copy of itself, C(t) = ⟨A(0)A(t)⟩. How quickly it decays shows how long the system keeps a trace of its earlier state. Integrals of such functions give relaxation times and transport coefficients, such as the diffusion coefficient through the Green-Kubo relations. (UNVERIFIED) autocorrelation function An analysis algorithm that estimates how strongly each atom scatters X-rays by approximating its atomic scattering factor with a sum of Gaussian functions, whose coefficients were fitted to quantum mechanical calculations of atomic electron density. It is used to compute theoretical X-ray scattering curves from a structure or a trajectory, so that a simulation can be compared with small-angle X-ray scattering (SAXS) experiments. (UNVERIFIED) cromer's method An enhanced sampling method that applies a bias which is updated as the simulation runs, driven by how evenly the chosen coordinate has been visited so far. The bias grows where the system has spent time and shrinks where it has not, which pushes it into unvisited regions. The free energy profile is recovered from the bias that was needed. (UNVERIFIED) accelerated weight histogram A simplified variable used to track the progress of a chemical reaction or a complex structural change. It reduces the motion of thousands of atoms into a single parameter that describes the transition from a starting state to a final state. Distinct from the object property MOLSIM_001036. Modeled here as a Class (the variable itself). reaction coordinate A computational technique used to anchor atoms to specific locations in 3D space during a simulation. It applies an artificial harmonic energy penalty proportional to the square of the displacement of an atom from a reference coordinate. For experts, these are heavily used during system equilibration (to relax solvent around a fixed solute) or in free energy calculations to prevent ligand drift, typically defined with a force constant k in units of kJ/(mol*nm^2) or kcal/mol. position restraints A specific interaction term used in molecular mechanics force fields to refine the vibrational behavior of bond angles. It introduces a harmonic potential acting on the distance between the 2 atoms at the ends of a bond angle (the 1-3 interaction), effectively placing a spring across the angle. It is a standard part of the CHARMM force field family, where it couples bond stretching and angle bending to better reproduce vibrational frequencies measured by spectroscopy. (UNVERIFIED) Urey-Bradley potential A non-bonded interaction parameter that lists the pairs of atoms whose ordinary electrostatic and van der Waals interactions are left out of the force calculation. It usually covers atoms joined by 1 or 2 covalent bonds, because their interaction is already described by the bond and angle terms. Force fields differ in how they treat atoms 3 bonds apart, which are often excluded and then added back as separate, scaled or specially parameterised pair interactions, so the rules must be kept when a topology is converted between programs. (UNVERIFIED) Defines the concept of the exclusion list/rule. exclusions A constraint that depends only on the positions of the particles, and possibly on time, and so reduces the number of degrees of freedom of the system. In molecular dynamics the usual examples are fixed bond lengths and the rigid geometry of a water molecule. Algorithms such as SHAKE, RATTLE and LINCS keep these constraints satisfied at every step by calculating the forces needed to hold them. (UNVERIFIED) holonomic constraint A particle mesh Ewald method that spreads the charges onto the grid with smooth B-spline functions, which allow the energy to be differentiated analytically to obtain the forces. Compared with the original particle mesh Ewald method, which uses Lagrange interpolation, this improves energy conservation and accuracy. It is the form of particle mesh Ewald used in programs such as AMBER and GROMACS, and its cost grows as N log N with the number of atoms N. (UNVERIFIED) SPME smooth particle mesh Ewald A numerical coefficient in a force field that quantifies the strength of the long-range attractive forces between two atoms due to induced dipole fluctuations. In the standard Lennard-Jones potential, this is the C6 term, proportional to the product of the interaction energy well depth and the sixth power of the collision diameter (4ϵσ^6). For experts, this parameter governs the asymptotic decay of the van der Waals potential and is often derived from atomic polarizabilities. (UNVERIFIED) dispersion interaction parameter A numerical coefficient in a force field that sets the strength of the short-range repulsion between 2 atoms, which stops their electron clouds from overlapping. In the Lennard-Jones potential, this is the C12 coefficient, equal to 4ϵσ^12, and the repulsion it describes arises from the Pauli exclusion principle. The 1/r^12 form is less physical than an exponential repulsion, and was chosen mainly because it is cheap to compute as the square of the 1/r^6 term. (UNVERIFIED) repulsion interaction parameter A simulation technique in which the positions of a selected group of atoms are held fixed throughout a run, because their positions and velocities are not updated in the chosen directions. Unlike position restraints, which pull atoms back towards a reference position and still let them move a little, frozen atoms do not move at all. It is used, for example, to model a rigid wall or surface, or to let solvent relax around a fixed protein during equilibration. (UNVERIFIED) atom freezing A modelling method that reduces a system's degrees of freedom by grouping atoms into single interaction sites according to chemical or structural rules. It is the act of building the model, as distinct from the resolution the resulting model has. The rule chosen for grouping determines what the model can still represent. (UNVERIFIED) coarse-graining A simulation method that adds a random force and a matching frictional drag to the equations of motion, imitating collisions with a surrounding fluid that is not itself simulated. The 2 additions are tied together so the system settles at the intended temperature. It also serves as a thermostat, which is why it is often used even when the solvent is present. (UNVERIFIED) stochastic dynamics A simulation method in which the system is deliberately driven away from equilibrium by an applied force, flow or gradient. Measuring how it responds gives transport properties such as viscosity and thermal conductivity directly, which an equilibrium simulation can only obtain indirectly from fluctuations. To obtain such a transport coefficient, the driving must be gentle enough that the response stays proportional to it. (UNVERIFIED) NEMD non-equilibrium molecular dynamics A computational technique used to force a protein out of its stable resting shape so it can explore new structures. It works by adding a localized repulsive energy hill, based on the system's principal components, to the energy minimum the molecule currently occupies. flooding potential Simulated tempering is an enhanced sampling technique that helps a system explore different configurations by allowing its temperature to fluctuate dynamically up and down during a single simulation. Unlike replica exchange, which uses parallel copies, this method uses a single trajectory that jumps between discrete temperature levels based on a probabilistic acceptance rule. It requires the pre-calculation or iterative determination of weighting factors for each temperature to ensuring uniform sampling and efficient barrier crossing. simulated tempering A classification code in a force field topology that selects which parameters apply to a bond, angle or rotation, based on the kinds of atoms involved. Assigning it is how a general parameter set is applied to a particular molecule. A wrong assignment gives a molecule the wrong stiffness or the wrong preferred shape, and nothing reports it. (UNVERIFIED) bonded type A salt bridge is a strong non-covalent interaction formed between a positively charged amino acid side chain and a negatively charged one. It combines two physical forces: an electrostatic attraction between the opposing charges and a hydrogen bond between specific atoms in the residues. For experts, these interactions (typically between Arg/Lys and Asp/Glu) contribute significantly to protein thermostability but have a complex energetic dependence on the local dielectric environment and pH. salt bridge A preparation process in which a solute is placed inside a box of solvent molecules, usually water, and any solvent that overlaps the solute is removed. The solvent box is normally pre-equilibrated, so the liquid already has the right density and structure before the solute enters it. Without solvent, either explicit or modelled implicitly, the molecule would be simulated in a vacuum, which can change its behaviour greatly. (UNVERIFIED) solvation The 2 ends of a biological polymer chain, which differ chemically from each other and from everything between them. In a protein one end carries a free amino group and the other a free carboxyl group, each charged at ordinary acidity. They must be assigned a chemical state before a simulation, because a standard residue template does not cover them. (UNVERIFIED) termini A configuration setting that instructs a molecular dynamics engine to resume a simulation exactly from where a previous run finished. It requires reading the full state of the system—including atomic positions, velocities, and sometimes forces—from a checkpoint or restart file to ensure a seamless trajectory. continuation state A sampling method that computes the chemical potential of a species by repeatedly placing a non-interacting trial particle at random positions in a simulation and recording the energy it would have. Averaging the Boltzmann factor of those energies gives the excess chemical potential directly. It fails for dense systems, where almost every insertion overlaps an existing particle. (UNVERIFIED) test particle insertion Gibbs sampling replica exchange is an advanced simulation method used to efficiently explore the different shapes and energy states of a molecule. Unlike standard replica exchange which attempts to swap temperatures between neighboring simulations, this method attempts to swap configurations among all running simulations simultaneously based on a global probability criterion. It is an advanced Markov chain Monte Carlo (MCMC) algorithm. It overcomes the problem of simulations getting trapped in low-energy "metastable" states by running multiple copies (replicas) of a system at different conditions. Gibbs sampling replica exchange A simulation method for polarizable force fields in which each polarizable atom is given a light charged particle attached to it by a spring. The particle shifts in response to the local electric field, which reproduces how an atom's electron cloud distorts. Treating polarization as an extra particle lets an ordinary integrator handle it. (UNVERIFIED) shell molecular dynamics A simulation method that forces a specific part of a molecular system, such as a protein subunit, to spin around a defined axis. An external potential is applied to a group of atoms to induce rotation at a constant angular velocity or under a constant torque while the rest of the system responds dynamically. enforced rotation simulation A barostat algorithm that controls pressure by giving the box volume its own equation of motion, coupled to Nosé-Hoover thermostat chains, following the equations of Martyna, Tuckerman, Tobias and Klein. Integrated with a velocity Verlet scheme, it samples the isothermal-isobaric (NPT) ensemble correctly apart from integration errors. In GROMACS it works only with velocity Verlet integrators, supports only isotropic scaling and cannot be combined with constraints. (UNVERIFIED) MTTK pressure coupling A non-bonded potential that combines an exponential repulsion with an attractive term falling off as the sixth power of distance, usually written V(r) = A exp(−Br) − C/r^6. The exponential describes the short-range repulsion between electron clouds more realistically than the r^−12 term of the Lennard-Jones potential. At very short distances the attractive term outgrows the repulsion and the energy falls without limit, so simulations need a safeguard to stop atoms collapsing onto each other. (UNVERIFIED) buckingham potential A computational model that uses machine learning algorithms to predict the potential energy and forces of a molecular system. It approximates the complex quantum mechanical energy landscape by training a neural network on a large dataset of reference atomic configurations and energies. For experts, these potentials (like ANI or DeepMD) offer ab initio accuracy at a computational cost closer to classical force fields, provided the simulation remains within the chemical space represented in the training data. (UNVERIFIED) neural network potential A multiscale simulation technique that couples a region described by a Neural Network Potential (NNP) with a larger environment described by a classical molecular mechanics (MM) force field. It allows for the accurate modeling of reactive centers or complex electronic effects using machine learning, while efficiently capturing long-range environmental effects with standard physics-based models. For experts, this method typically employs electrostatic or mechanical embedding schemes similar to QM/MM to handle the boundary interactions between the NNP-active region and the classical surroundings. (UNVERIFIED) hybrid NNP/MM simulations hybrid NNP/MM simulation A simulation technique that allows a user to visualize and physically manipulate a running molecular simulation in real-time. Through a specialized interface or haptic device, the user applies external forces to specific atoms while the simulation engine calculates the system's response on the fly. For experts, IMD uses a client-server architecture where the visualization client sends force vectors to the simulation server, which integrates these as time-dependent external potentials, useful for guiding docking or unfolding pathways. (UNVERIFIED) interactive molecular dynamics An integration algorithm that updates velocities at half-integer time steps and positions at whole time steps, so that the 2 updates alternate. It is time-reversible and symplectic, which means it keeps the total energy close to its starting value over long simulations without a steady drift. Because positions and velocities are not known at the same moment, quantities such as the kinetic energy at a whole step are estimated from the velocities of the 2 adjacent half steps. (UNVERIFIED) leapfrog integrator A numerical algorithm for integrating Newton's equations of motion that updates atomic positions and velocities at the same time step. It gives the same trajectory as the leapfrog method, but provides positions and velocities at the same moment, which simplifies the calculation of the pressure and the coupling of thermostats and barostats. It is symplectic and time-reversible, and is the standard integrator in many molecular dynamics programs. (UNVERIFIED) velocity verlet integrator A numerical algorithm used to simulate the motion of particles in a fluid where friction dominates over inertia. It integrates the overdamped Langevin equation, assuming that the momenta relax instantaneously, so particle positions are updated based on forces and a random displacement derived from the diffusion coefficient. For experts, this integrator allows for much larger time steps than standard MD by ignoring fast inertial motions, making it ideal for simulating diffusion-controlled processes or coarse-grained systems. (UNVERIFIED) brownian dynamics integrator A simulation method that models the movement of molecules as a random walk driven by thermal collisions and damped by solvent viscosity. Unlike standard molecular dynamics, it ignores the mass and inertia of the particles, assuming that friction brings them to a terminal velocity instantly in response to forces. For experts, this is the limit of Langevin dynamics where the friction coefficient is very large (overdamped regime), often used for simulating large biomolecules or coarse-grained models where solvent is treated implicitly. (UNVERIFIED) brownian dynamics A numerical algorithm that incorporates random noise and frictional drag directly into the equations of motion to mimic the effect of a solvent or a heat bath. It integrates the Langevin equation, or a similar stochastic differential equation, so that it propagates the system and controls its temperature at the same time. With a small friction it acts as a thermostat that works on each particle separately and samples the canonical ensemble. (UNVERIFIED) stochastic dynamics integrator A triclinic box is a geometric shape for a periodic simulation cell defined by three vectors of arbitrary length and mutual angles. This flexibility allows it to represent any crystal lattice system or to optimize the solvent volume around a solute by fitting the box shape more closely than a standard cube. triclinic box A barostat algorithm that rescales the box and the particle coordinates so that the pressure relaxes exponentially towards its target, as in the Berendsen barostat, with an added random term. The random term is what makes it produce the correct distribution of volumes rather than merely the correct average. It reaches equilibrium quickly like the simple rescaling schemes while sampling the isothermal-isobaric ensemble correctly, so it can be used for both equilibration and production. (UNVERIFIED) baro_stochastic=1 pcoupl=C-rescale stochastic cell rescaling barostat baro_stochastic=1 pcoupl=C-rescale An integration parameter that names the algorithm used to advance the coordinates and velocities of the system from one step to the next, such as leap-frog or velocity Verlet. In some programs the same setting also selects stochastic dynamics, such as a Langevin integrator, or a method that is not an integration over time at all, such as steepest descent energy minimization. The choice affects the numerical stability of a run and, for stochastic integrators, how the temperature is controlled. (UNVERIFIED) integrator type An output control parameter that specifies how often the atomic coordinates of the system are written to the trajectory file during a run. It is usually given as the number of integration time steps between 2 saved frames. A shorter interval gives finer time resolution for later analysis, at the cost of larger files and more time spent writing them. (UNVERIFIED) trajectory output interval A temperature control parameter that divides the system into groups of atoms, such as the protein and the solvent, each coupled to its own thermostat. Coupling the groups separately can prevent one part of the system from becoming hotter than another, a known problem when solvent and solute are coupled as 1 group. Each group must be large enough to justify its own thermostat, so the system is not split into many small groups. (UNVERIFIED) temperature coupling groups A non-bonded interaction parameter that selects how the simulation program builds its neighbor lists and applies cutoffs to non-bonded interactions, either per atom with a buffered list or per charge group. In the buffered scheme, called Verlet in GROMACS, the list also holds pairs slightly beyond the cutoff, and the size of this buffer is set automatically from a tolerance for the error it may cause. The older charge-group scheme, called group in GROMACS, is no longer supported. (UNVERIFIED) cutoff scheme A non-bonded interaction parameter giving the distance within which pairs of atoms are stored in the neighbor list. It is equal to or larger than the interaction cutoff, and the extra margin lets the list stay valid for several steps while atoms move. Pairs farther apart than this distance are not considered again until the list is next rebuilt. (UNVERIFIED) neighbor list cutoff A system setup parameter that says whether starting velocities are assigned to the atoms by drawing them at random from the Maxwell-Boltzmann distribution for a chosen temperature. A structure alone carries no velocities, so they must be generated before dynamics can start from a new structure. When a run continues from an earlier one, the velocities are normally read from the restart file instead and are not generated again. (UNVERIFIED) velocity generation A function giving the density of a chosen atom or molecule at each point in the space around a reference molecule, relative to its average density in the bulk. Unlike a radial distribution function, which averages over all directions and returns a curve, it keeps the direction and returns a 3-dimensional map. That is what lets specific sites be seen, such as a water molecule that sits in the same pocket throughout a simulation. (UNVERIFIED) SDF spatial distribution function A binary file format used by the GROMACS simulation package to store time-series data of system energies and thermodynamic properties. It records values such as potential energy components, temperature, pressure, and box dimensions at specified intervals in a portable, machine-independent format. For experts, this file (extension .edr) is accessed via the gmx energy tool to extract statistics or plotting data, and its binary nature ensures no loss of precision compared to text logs. (UNVERIFIED) GROMACS portable energy format A text data format used by GROMACS to define named groups of atoms within a molecular system, each group given as a name in square brackets followed by the indices of its atoms. It lets programs apply an action, such as temperature coupling, a restraint or an analysis, to just those atoms. Files in this format have the extension .ndx and are made with gmx make_ndx or gmx select when the default groups, such as Protein, are not enough. (UNVERIFIED) GROMACS index format A binary file format used by GROMACS that creates a self-contained description of a simulation ready to be executed. It combines the atomic coordinates, velocities, molecular topology, and all simulation control parameters into a single, portable machine-readable file. For experts, this file (extension .tpr) is generated by the gmx grompp preprocessor and ensures consistency by validating that the structure matches the topology and parameters before the simulation engine is launched. (UNVERIFIED) GROMACS portable input format An atom ID is usually a number, that uniquely defines a given atom in a molecular structure. It is typically used in molecular topology descriptions to uniquely identify an atom from the set of atoms defining the system. atom ID JSmol is an open-source, HTML5/JavaScript-based molecular viewer that allows for the interactive 3D visualization of chemical structures within web browsers without requiring Java or other plugins. It is the modern, JavaScript implementation of the popular Jmol applet, widely used in chemical education, databases, and web-based cheminformatics tools. JSmol A visualization tool designed specifically to render and interactively display three-dimensional models of molecular structures, complexes, and dynamic simulation trajectories. These tools are used to visually inspect atomic coordinates, molecular surfaces, and spatial relationships. molecular structure visualization tool A visualization tool designed to generate 2D or 3D graphs, charts, and plots from numerical data files. data visualization tool data plotting tool A simulation program that performs both molecular dynamics and Monte Carlo sampling of biological macromolecules. It is used particularly for intrinsically disordered proteins, where no single structure exists and broad sampling matters more than speed. Its Monte Carlo moves can change a chain's shape in ways dynamics would take far longer to reach. (UNVERIFIED) CAMPARI A programming interface, developed by AMD, that lets the same source code run on graphics processors from more than one manufacturer. It closely mirrors the interface used for NVIDIA hardware, so existing code can be converted with little change. Simulation engines use it so that 1 version of their graphics processor code serves both AMD and NVIDIA hardware. (UNVERIFIED) HIP A standard C++ programming layer that lets code for processors of different kinds, such as central processors and graphics processors from several manufacturers, be written as ordinary C++ in a single source file. The parts that run on the accelerator are compiled from the same file as the rest of the program, so no separate device language is needed. GROMACS, for example, uses it to run on graphics processors from more than one manufacturer. (UNVERIFIED) SYCL An open standard for writing programs that run across processors of different kinds, including central processors, graphics processors and specialised accelerators. One program can then use whatever hardware is present. Its portability comes at some cost in speed against an interface written for one manufacturer. (UNVERIFIED) open computing language OpenCL An open-source implementation of the message passing standard, formed by merging several earlier projects and designed to support many kinds of network hardware. It is one of the most widely used implementations, alongside MPICH. Which implementation a simulation engine was built against can affect its performance. (UNVERIFIED) OpenMPI An open-source implementation of the message passing standard, used directly and also as the basis several hardware vendors build their own versions on. It is one of the most widely used implementations, alongside Open MPI. Its performance depends heavily on the network hardware it is configured for. (UNVERIFIED) MPICH AutoDock is a suite of automated docking tools designed to predict how small molecules bind to a receptor of known 3D structure. AutoDock (e.g., AutoDock 4) utilizes a Lamarckian genetic algorithm and a grid-based empirical scoring function. AutoDock GOLD (Genetic Optimisation for Ligand Docking) is a highly regarded commercial docking program developed by the Cambridge Crystallographic Data Centre (CCDC). It uses a genetic algorithm to explore ligand conformational space and offers multiple scoring functions. It is frequently used in pharmaceutical pipelines to generate highly accurate protein-ligand complex structures for subsequent MD refinement. Genetic Optimisation for Ligand Docking GOLD Glide (Grid-based Ligand Docking with Energetics) is a commercial molecular docking program within the Schrödinger software suite. It uses a systematic, hierarchical search of positional, orientational, and conformational space followed by evaluation with an empirical scoring function. It is heavily utilized in industry for virtual screening and preparing complexes for Desmond MD simulations. Grid-based Ligand Docking with Energetics Glide A molecular docking program that predicts how a small molecule will bind to a protein or a nucleic acid, and scores how well it fits. It searches for good positions using a mix of random and directed steps, which lets it cover many possibilities without testing them all. It grew out of the earlier commercial program RiboDock and was released as open source in 2012. (UNVERIFIED) rDock An enhanced sampling algorithm used in molecular dynamics to calculate free energy profiles and accelerate the sampling of complex molecular systems. By continuously applying an on-the-fly biasing force, ABF allows systems to escape kinetic traps and evenly explore reaction pathways. ABF algorithm adaptive biasing force algorithm Wang-Landau algorithm simulations are a class of flat-histogram Monte Carlo methods designed to directly estimate the density of states (DOS) of a physical or mathematical system. Unlike traditional methods dependent on temperature, it generates a random walk in energy space that allows you to calculate thermodynamic properties across all temperatures. Wang-Landau algorithm A specific algorithmic extension of metadynamics where the height of the added Gaussian biasing potentials is gradually decreased over time based on the amount of bias already accumulated at that location. This ensures that the biasing potential converges smoothly to the true free energy surface, avoiding the over-filling of energy basins. well-tempered metadynamics algorithm An algorithm that applies a time-dependent, usually harmonic, guiding potential to specific atoms or collective variables. It calculates the necessary biasing forces to force the system to evolve along a predefined reaction coordinate, simulating non-equilibrium processes like ligand unbinding or protein unfolding. SMD algorithm steered molecular dynamics algorithm An enhanced sampling simulation method and a specific variant of metadynamics where multiple independent simulations (walkers) are run in parallel. All walkers simultaneously explore the conformational space and deposit Gaussian bias potentials into a single, shared history-dependent bias grid. This approach significantly accelerates the reconstruction of complex free energy surfaces while maintaining the exact theoretical accuracy of standard metadynamics. MWMD multiple-walkers metadynamics The PDBQT format is an extension of the standard PDB format specifically designed for use with the AutoDock and AutoDock Vina molecular docking suites. In addition to atomic coordinates, it stores calculated partial charges ('Q') and AutoDock-specific atom types ('T'), making it a structural format that intrinsically carries the predictive interaction parameters required for, and generated by, docking algorithms. PDBQT format A text data format recording the predicted acidity of each titratable group in a protein, together with the structural reasons for each prediction. It lists the contributions separately, such as burial from solvent and nearby charges, so a prediction can be checked rather than taken on trust. It is normally written with the extension .pka. (UNVERIFIED) pka format PROPKA output format A docking log format is a text-based or tabular file generated by molecular docking software (such as AutoDock Vina, Glide, or GOLD) to record the numerical results of a virtual screening or pose prediction run. It typically contains ranked lists of predicted binding poses, estimated binding affinities or docking scores, and clustering metrics like RMSD, serving as the primary predictive output for analyzing binding without needing to parse the 3D structure files. docking log format A data format, usually written as JSON, that stores the predicted aligned error matrix from a deep learning protein structure prediction. Entry X,Y gives the expected error in the position of residue X when the predicted and true structures are superposed on residue Y. It is how the confidence in relative domain orientation and in predicted interfaces is judged, which a per-residue confidence score cannot show. (UNVERIFIED) Predicted Aligned Error matrix format PAE matrix format The Maestro format (typically .mae or .maegz) is the native, proprietary text-based format developed by Schrödinger software suite, used to store molecular structures along with associated properties. mae format maegz format Maestro format A sampling algorithm that proposes a random change to a system, always accepts it if the energy falls, and otherwise accepts it with probability exp(-ΔE/kT), where ΔE is the energy rise, k the Boltzmann constant and T the temperature. Repeating this generates configurations in the right proportions for the temperature, without ever computing a force. It samples equilibrium properties correctly but says nothing about how fast anything happens. (UNVERIFIED) metropolis Monte Carlo A simulation engine, the program in the GROMACS suite that performs the simulation itself. It reads a single pre-compiled input file holding the system, the parameters and the starting state, and writes the trajectory. Every other GROMACS program prepares its input or analyses its output. (UNVERIFIED) mdrun A data format that specifies the control variables, algorithmic settings, and thermodynamic conditions required to execute a computational simulation run. It dictates exactly how the simulation engine will propagate the system over time, defining parameters such as the integration algorithm, time step, total duration, non-bonded cutoff distances, and the specific thermostat and barostat algorithms used for temperature and pressure coupling. For experts, files in this format (such as GROMACS .mdp or AMBER mdin) define the physical ensemble and numerical methods of the computational experiment, acting as the primary instruction set for the core molecular dynamics engine. (UNVERIFIED) simulation parameter format A data format that stores the numerical parameters defining the potential energy function of a molecular mechanics force field, given for atom types rather than for individual atoms. It holds equilibrium values and force constants for bonds, angles and dihedrals, and van der Waals parameters for non-bonded interactions, while partial charges are usually supplied separately with the residue definitions. Examples are the AMBER .dat and frcmod files and the CHARMM parameter file. (UNVERIFIED) force field parameter format A simulation environment represents the overall physical medium or surrounding context in which the primary molecular system is embedded during a computational study. It defines the macroscopic nature of the system's surroundings, such as whether the molecule is in a vacuum, dissolved in a liquid, or embedded in a lipid bilayer. simulation environment An environment in which a molecule is simulated on its own, with no solvent or other surroundings, as in the gas phase. Without solvent to screen charges and to compete for hydrogen bonds, the charged and polar groups of a flexible molecule tend to interact with each other, so its shape can differ strongly from that in solution. It is used for parameterisation, for comparison with gas-phase measurements such as mass spectrometry, and for quick energy minimisations. (UNVERIFIED) vacuum environment A solvated environment is a simulation context in which the primary solute molecule is completely surrounded by a liquid medium composed of solvent molecules. This environment is characterized by bulk liquid properties such as density, viscosity, and a specific dielectric constant. Simulating molecules in a solvated environment is essential for capturing the thermodynamic and structural effects of solvent interactions, which heavily dictate biomolecular folding and binding. solvated environment A solvated environment in which the liquid is water, alone or nearly so. It is the setting almost all biological molecules exist in, so it is the default choice for a biomolecular simulation. The behaviour of water near a surface differs enough from bulk water that the amount included matters. (UNVERIFIED) aqueous environment A solvated environment in which the liquid is something other than water, such as methanol, chloroform or octanol. These solvents differ from water in polarity and in their ability to form hydrogen bonds, so a molecule adopts different shapes in them. Simulations in octanol are used to predict how a drug will partition between water and fat. (UNVERIFIED) non-aqueous environment A membrane environment is a highly anisotropic, heterogeneous simulation context representing a cellular membrane, typically composed of a self-assembled lipid bilayer separating two aqueous compartments. It features a hydrophobic core and polar interfacial regions. Modeling a membrane environment is essential for simulating the structure, dynamics, and transport mechanisms of transmembrane proteins, ion channels, and membrane-active peptides or drugs. membrane environment An environment in which molecules are packed in a regular repeating arrangement fixed by a crystallographic space group, as in a solid. Neighbouring molecules constrain each other, so shapes seen here may not be those the molecule takes in solution. Simulations in this environment are used to study crystal packing and to compare a model with the crystal it was measured in. (UNVERIFIED) crystal lattice environment An interfacial environment is a simulation context characterized by the boundary or interface between two distinct physical phases, such as a liquid-vapor (water-air) interface or a liquid-solid interface. The physical properties of the system change drastically across this boundary. interfacial environment A dataset consisting of a collection of distinct molecular structures or conformations. Unlike a simulation trajectory, it does not inherently possess a chronological order, but rather represents a sampling of the structurally accessible states of a molecule (e.g., an NMR structural ensemble, or representative frames extracted via clustering). (UNVERIFIED) structural ensemble conformational ensemble A static computational method that calculates the potential energy of a molecule based on its 3D geometry, treating atoms as classical particles and chemical bonds as flexible springs. It uses an empirical force field to find the lowest-energy conformation (equilibrium geometry). MM molecular mechanics method molecular mechanics A method that estimates the binding affinity of 2 molecules, such as a drug and a protein, as a weighted sum of countable features of the complex, such as hydrogen bonds, buried surface and rotatable bonds, with the weights fitted to measured affinities. Because these terms are quick to compute, it is fast enough to score millions of candidate molecules, which is why docking programs use it. (UNVERIFIED) empirical scoring function empirical scoring method A modelling method that represents the effect of a surrounding liquid as a continuous medium instead of as individual molecules. It computes solvation free energy from the shape and charges of the solute using parameters such as atomic radii and a surface tension coefficient. It removes most of the atoms from a simulation, which makes it far faster but blind to any effect that depends on a specific solvent molecule. (UNVERIFIED) continuum solvation model implicit solvent method A modelling method that represents a macromolecule as nodes, usually 1 per residue, joined by simple springs between neighbours within a cutoff distance. It discards chemistry entirely and keeps only shape and connectivity, which is enough to reproduce large-scale collective motion. It is cheap enough to apply to assemblies far too large to simulate. (UNVERIFIED) ENM elastic network modeling A modelling method that predicts a molecule's biological effect from numbers computed about its chemical structure. It fits a statistical relationship on compounds whose activity is known, then applies it to compounds whose activity is not. It predicts well only for molecules resembling the training set, which is why the range it was fitted over must be reported with it. (UNVERIFIED) QSAR quantitative structure-activity relationship modeling A computational method that predicts the energy and forces of a system using a model trained on quantum mechanical calculations, instead of a fitted analytical force field. It aims for quantum accuracy at a cost close to a force field's. It is reliable only for systems resembling those it was trained on, which is its central limitation. (UNVERIFIED) machine-learned potential method A protein secondary structure change occurs when the local spatial conformation of a polypeptide backbone, primarily alpha-helices, beta-sheets, turns, and random coils, transitions from one shape to another. These changes are stabilized or disrupted by alterations in backbone hydrogen bonding. secondary structure change The physical process by which a linear chain of amino acids transforms into a functional, three-dimensional structure. This 3D shape is essential for the protein to carry out its specific biological tasks. folding Protein unfolding transitions a native, three-dimensional functional protein back into an unordered or partially structured polypeptide chain. It is often simulated instead of folding, because it can be driven to happen fast enough to be observed in a reasonable timescale. unfolding A collective motion in which a whole folded region of a protein shifts or rotates relative to the rest. It usually happens by bending or twisting at a hinge joining 2 more rigid regions. It is the kind of motion that opens and closes a binding site. (UNVERIFIED) domain motion A structural change in a protein or multi-subunit complex triggered when a small molecule (an effector or modulator) binds to a site separate from the active site. allosteric transition An allosteric transition set off by a small molecule binding at one site, which changes behaviour at a distant one. The effect travels through connected motions in the structure rather than through direct contact. It is how a drug can switch a protein's activity without touching its active site. (UNVERIFIED) ligand-induced allosteric transition An allosteric transition in which the signal is carried by a change in flexibility rather than by a change in average shape. A perturbation at one site alters how much a distant site fluctuates, which changes its affinity without moving it. Because the average structure barely changes, crystallography alone cannot detect it. (UNVERIFIED) entropy-driven allosteric transition A conformational rearrangement in which the five-membered sugar ring of a nucleotide shifts between different non-planar shapes. In nucleic acids, this commonly involves transitions between North (C3′-endo-like) and South (C2′-endo-like) conformations, affecting local backbone geometry. Correctly reproducing these populations is an important test of nucleic-acid force fields and helps distinguish A- and B-form helices. sugar ring flip sugar ring pucker transition A physical constituent of a simulated molecular system that is treated as a distinct unit with its own properties and spatial coordinates. It covers atoms, ions, residues, polymer chains and whole molecules. (UNVERIFIED) Top-level parent for the structural components of a simulated system. Kept as a MOLSIM class rather than reusing CHEBI:23367 molecular entity, because this branch spans 2 ChEBI branches that do not meet: atom and molecule are molecular entities in ChEBI, while residue and chain sit under CHEBI:24433 group, a sibling of molecular entity. Grouping all 4 suits the description of a coordinate file, where they are 4 levels of one containment hierarchy. Renamed from 'molecular entity' on 2026-08-21 so that the primary label is unique across the ontologies MOLSIM is released alongside; the former name is kept as an exact synonym. molecular entity molecular structural component A group of atoms that a structure file handles as 1 named and numbered unit, normally a single building block of a polymer chain, such as an amino acid in a protein or a nucleotide in a nucleic acid. Each water molecule, ion or small ligand is usually stored as a residue of its own as well. Naming the unit lets a file or an analysis refer to "residue 45" rather than to each of its atoms. (UNVERIFIED) residue The smallest unit of a chemical element, and the smallest particle a classical molecular simulation tracks. In such a simulation it is represented as a point carrying a mass, a partial electric charge and a radius, all taken from the force field rather than computed. Everything else in a simulated system is built from these points and the bonds between them. (UNVERIFIED) atom A continuous run of residues joined by covalent bonds and labelled with a single identifier in a structure file, such as a single strand of DNA or a single polypeptide. The chain identifier is what separates the subunits of a complex from one another. A wrong identifier can make software treat 2 separate molecules as 1, or 1 molecule as 2. (UNVERIFIED) chain A molecular structural component made of atoms joined by covalent bonds into a single continuous structure that shares no covalent bond with anything else in the system, or a single monatomic ion. It ranges from a single solvent molecule to a whole protein. Simulation software needs to know where each molecule begins and ends, for example to keep molecules whole when they cross the edge of a periodic box and to compute how far each one moves. (UNVERIFIED) molecule molecular unit A pocket detection tool a computational algorithm that identifies and characterizes transient or persistent cavities on the surface of a biomolecular structure from a trajectory or static coordinate set, typically by analyzing atomic coordinates, solvent accessibility, or geometric properties, to predict potential ligand-binding sites and assess their dynamic accessibility and druggability. cavity detection tool pocket detection tool fpocket is a fast, open-source algorithm designed to predict and analyze potential binding pockets (cavities) on proteins and other biomolecules. It is used in structural bioinformatics for applications like drug discovery, protein function annotation, and the analysis of molecular dynamics trajectories. fpocket Curves+ is a software program for the structural analysis of DNA and RNA three-dimensional structures. It calculates a comprehensive set of geometric parameters describing nucleic acid conformation, including base-pair, base-pair step, backbone, groove, and helical axis parameters, using a reference frame attached to each base. These descriptors provide a detailed characterization of nucleic acid structure and flexibility. Curves+ An execution that resumes a previously interrupted or completed molecular dynamics run by reading saved coordinates, velocities, and simulation state from a restart file. It allows a long simulation to be continued across multiple sessions without loss of continuity. (UNVERIFIED) restart simulation process A data processing tool that reads simulation input, output, or trajectory files and extracts descriptive metadata such as system composition, run settings, and provenance information. It supports cataloguing simulation datasets and describing them so that they are findable, accessible, interoperable and reusable (FAIR). (UNVERIFIED) metadata extraction tool A build of the AMBER simulation engine that runs on NVIDIA graphics processors. It performs the whole simulation on the graphics card, so data need not be moved back and forth each step, which is where the speed comes from. It supports fewer methods than the processor build, so which was used should be recorded. (UNVERIFIED) PMEMD.cuda A validation analysis that checks recorded or extracted metadata for completeness, correctness, and internal consistency against expected schemas or physical constraints. It helps ensure that a simulation dataset is reliably described before it is shared or archived. (UNVERIFIED) metadata validation process A trajectory post-processing that translates atomic coordinates back into the primary periodic unit cell. coordinate wrapping process A restrained simulation method that biases the dynamics using experimental nuclear magnetic resonance observables, such as interatomic distance restraints from nuclear Overhauser effects or torsion-angle restraints from scalar couplings. It guides the simulated structure toward conformations consistent with the measurements. (UNVERIFIED) NMR restrained simulation A restrained subset dynamics in which only a selected group of atoms is allowed to move, while the coordinates of all other atoms are held fixed throughout the run. Freezing atoms this way gives no significant speed advantage, and restraining atoms close to their starting positions is now usually preferred. (UNVERIFIED) belly dynamics An enhanced sampling parameter that specifies how often exchanges between neighboring replicas are attempted during a replica exchange simulation. It controls the rate at which configurations are swapped between conditions such as different temperatures or Hamiltonian λ values. replica exchange attempt frequency An integration parameter that specifies how frequently the net translational motion of the system's center of mass is removed during a run. Periodic removal prevents the gradual accumulation of spurious overall drift of the simulated system. (UNVERIFIED) center of mass removal interval An integration parameter that specifies how many short inner integration steps are taken per long outer step in a multiple-timestep integration scheme. It sets how often fast-varying forces are evaluated relative to slow-varying forces to improve computational efficiency. (UNVERIFIED) multi-timestep inner loop count A simulation parameter that governs how frequently and in what form simulation data are written to output files during a run. It determines the temporal resolution and volume of the recorded energies, coordinates, and diagnostic information. (UNVERIFIED) output control parameter An output control parameter that specifies how often energy and related thermodynamic quantities are written to the output or log during a run. Smaller intervals give finer-grained monitoring at the cost of larger output files. (UNVERIFIED) energy output interval An output control parameter that specifies how often a restart file containing the current coordinates, velocities, and state is written during a run. It determines how much progress can be recovered if the run is interrupted. (UNVERIFIED) restart output interval An output control parameter that specifies how often diagnostic and status information is written to the simulation log during a run. It controls the verbosity and temporal resolution of the recorded run history. (UNVERIFIED) log output interval A simulation state that represents the elapsed model time reached by the system at a given point in a molecular dynamics run. It records how far the simulation has advanced along the physical time coordinate. simulation state time A thermodynamic property that represents the mass density of the simulated system at the end of a run or equilibration stage. It is commonly compared against experimental density to assess whether the system has equilibrated correctly. final density A computed observable that quantifies the computational cost or efficiency of a simulation run rather than a physical property of the system. Examples include elapsed real time and the amount of simulated time produced per unit of computing time. (UNVERIFIED) computational performance metric A computational performance metric that measures the total real-world elapsed time taken to complete a run from start to finish. It reflects the actual time a user waits, including all computation and input or output. (UNVERIFIED) total wallclock time A computational performance metric that measures the cumulative processor time consumed by a run, summed across all active compute cores. It reflects total computational effort rather than elapsed real time. (UNVERIFIED) total cpu time A computational performance metric that measures the real-world elapsed time spent preparing a run before the production phase begins, such as reading inputs and building data structures. It isolates initialization overhead from the main computation. setup wallclock time A computational performance metric that measures the real-world elapsed time spent in the production phase of a simulation run, during which the trajectory of interest is generated. It excludes setup and finalization overhead. production wallclock time A computational performance metric that represents the mean real-world time required to advance the simulation by one integration step. It is a convenient measure of per-step computational efficiency. (UNVERIFIED) average integration step time A computational performance metric that expresses how much simulated model time is produced per unit of real-world computing time, commonly reported as nanoseconds per day. It is a standard measure for comparing the practical speed of simulations. (UNVERIFIED) simulation throughput A plan specification that describes a sequential, multi-stage computational workflow, such as minimization, heating, equilibration, and production, applied to a molecular system. It organizes the individual stages and their ordering into a single reproducible procedure. (UNVERIFIED) simulation protocol A molecular dynamics process that constitutes one discrete, chronological phase within a larger simulation run, characterized by its own role and its own input and output files. Examples include a heating stage or a production stage. (UNVERIFIED) simulation stage A simulation parameter that specifies how another parameter, such as target temperature, restraint force constant, or non-bonded cutoff, changes over the course of a run, whether across simulated time or across integration steps. It captures ramping and stepwise schedules applied during a stage. simulation parameter schedule A validation analysis that evaluates whether 2 consecutive simulation stages join without an unexpected break, by comparing the end time of one stage with the start time of the next. It flags gaps that may indicate missing segments or unit errors. (UNVERIFIED) simulation continuity check A validation analysis that checks whether metadata extracted from several distinct files describing the same stage, such as topology, input, output, and trajectory files, agree with one another. It surfaces inconsistencies such as mismatched atom counts or timesteps. (UNVERIFIED) cross-source validation An information content entity that records a failure, exception, or warning encountered while reading or extracting metadata from a simulation data file. It preserves the reason a file could not be fully processed. (UNVERIFIED) file parsing error A data format that designates a specific standardized schema variant that a file adheres to within a broader container format, such as the AMBER conventions used inside a NetCDF file. It identifies the precise rules for interpreting the file's contents. (UNVERIFIED) file format convention An information content entity that assigns a molecular system to a descriptive category, for example by its overall composition, by the biological function of its main protein, or by the organism strain or histone variant it represents. It gives a short, human-readable summary of what is being simulated, so that simulations of the same kind can be grouped and found. (UNVERIFIED) system classification label An atom selection specification that defines a subset of atoms within a molecular system using a formal string or logical expression following the format of a specific software. atom selection mask A simulation state time that represents the average amount of simulated time elapsed between consecutive saved frames of a trajectory. It reflects how finely the trajectory samples the simulated time coordinate. (UNVERIFIED) trajectory frame interval A simulation state time that represents the difference between the end time of one simulation stage and the start time of the next. A nonzero gap indicates a break in the continuous simulated-time coordinate across stages. (UNVERIFIED) inter-stage time gap An analysis algorithm that identifies which atoms in a molecular system are hydrogens, for the purpose of applying or detecting hydrogen mass repartitioning. Different strategies rely on different atomic attributes, such as atomic number or atom name. (UNVERIFIED) hydrogen detection strategy A thermodynamic property that represents the mass density of the simulated system at the start of a run, computed from the topology mass and the initial periodic box volume. It provides a baseline for comparison against the equilibrated density reached during a constant-pressure run. initial density A cross-source validation that checks whether the atom count agrees across all files describing a system, such as topology, trajectory, etc. A mismatch indicates the files do not describe the same system. atom count consistency check A cross-source validation that checks whether periodic box information is consistently present or absent across the files describing a stage. It flags cases where, for example, a topology declares a box but a coordinate file does not. (UNVERIFIED) box consistency check A cross-source validation that checks whether step counts, timesteps, and total durations agree across the input, output, and trajectory files of a stage. It verifies that the recorded trajectory length matches the configured number of steps times the timestep. (UNVERIFIED) timing consistency check A cross-source validation that checks whether the coordinate write frequency configured in the input matches the frequency reflected in the output and trajectory. It confirms that sampling settings were applied as specified. (UNVERIFIED) sampling consistency check A polypeptide domain that occurs in a protein. It is a distinct, self-contained part of a protein that folds up on its own and usually carries out a particular job. Large proteins are often built from several domains. protein domain A protein domain whose function is to attach to a matching partner molecule called a receptor. receptor-binding domain A protein domain that contains the active site, the pocket where an enzyme carries out its chemical reaction. It is the working part of an enzyme. catalytic domain A system classification label that records the biological job of the main protein in a simulated system. Proteins do most of the work in living cells, and this label groups simulations by the kind of job their protein does. (UNVERIFIED) protein functional class label A protein functional class label for a system whose main protein is an enzyme, a protein that speeds up a chemical reaction without being used up itself. (UNVERIFIED) enzyme classification label A protein functional class label for a system whose main protein is a transport protein, one that moves ions or molecules from place to place, for example carrying them across a cell membrane. (UNVERIFIED) transport protein classification label A protein functional class label for a system whose main protein is a structural protein, one that gives shape, stiffness or mechanical support to cells and tissues. (UNVERIFIED) structural protein classification label A protein functional class label for a system whose main protein is a regulatory protein, one that switches biological processes on or off, or adjusts how fast they run. (UNVERIFIED) regulatory protein classification label A protein functional class label for a system whose main protein is a signaling protein, one that passes messages within a cell or between cells so that the cell can respond to its surroundings. (UNVERIFIED) signaling protein classification label A protein functional class label for a system whose main protein is a motor protein, one that turns chemical energy into movement, for example by moving along protein filaments to carry cargo or to generate force. (UNVERIFIED) motor protein classification label A protein functional class label for a system whose main protein is an immune protein, one that helps recognise or fight off germs and other foreign material. Antibodies are a familiar example. (UNVERIFIED) immune protein classification label A protein functional class label for a system whose main protein is a storage protein, one that holds onto useful substances such as iron or amino acids until they are needed. (UNVERIFIED) storage protein classification label A system setup parameter that gives the three angles (commonly α, β, γ) between the edges of the simulation box. These are needed to describe boxes that are not simple rectangular blocks, for example triclinic or truncated-octahedral boxes. (UNVERIFIED) periodic box angles An identifier that encodes the structure of a molecule as a short line of text, writing atoms as element symbols and showing bonds, branches and rings with punctuation and digits. It lets a small molecule be written, searched and shared as plain text, and complements the InChI code. The name stands for Simplified Molecular-Input Line-Entry System. (UNVERIFIED) SMILES code A software development and management tool that translates a program's source code into a runnable program. Which compiler was used can affect a simulation's speed and, occasionally, its numerical results. (UNVERIFIED) compiler An atom selection specification that defines a named group of atoms as an explicit list of atom indices, so the group can be referred to by name during setup or analysis. Groups of this kind are what a GROMACS index (.ndx) file holds, for example Protein or Water. (UNVERIFIED) atom index group A molecular dynamics method in which the forces between atoms come from a classical (empirical) force field rather than from quantum-mechanical calculations. It is the most common and computationally affordable form of molecular dynamics. (UNVERIFIED) classical molecular dynamics An information content entity recording how the atoms and residues of a simulated system correspond to those of the experimental structure it was built from. The 2 rarely match: residues are renumbered, missing loops are built in, and hydrogens are added. Without this record, a simulation result cannot be related back to the numbering used in the literature. (UNVERIFIED) source structure residue mapping A structural property that describes how many copies of a protein chain join together to form a working unit. Many proteins function only when several copies assemble into one complex. (UNVERIFIED) oligomeric state An oligomeric state in which the protein works on its own as a single chain, not joined to copies of itself. A protein simulated as a single chain when it is really a pair will behave wrongly, because the surface that should be buried against its partner is exposed to water instead. The state is usually known from the experimental structure or from measurements in solution. (UNVERIFIED) monomeric state An oligomeric state in which 2 copies of a protein chain join together to form the working unit. The 2 copies bury a surface against each other that would otherwise face the solvent, so simulating only a single copy gives a different answer. Many enzymes and many transcription factors work this way. (UNVERIFIED) dimeric state An oligomeric state in which 3 copies of a protein chain join together to form the working unit. Such an assembly usually has 3-fold symmetry, so its subunits are equivalent and the whole can be built from 1. Many viral spike proteins and several membrane channels take this form. (UNVERIFIED) trimeric state An oligomeric state in which 4 copies of a protein chain join together to form the working unit. The copies need not be identical: haemoglobin, for example, is built from 2 copies each of 2 related chains. In haemoglobin the 4 subunits influence one another, so oxygen binding at 1 subunit makes binding at the others easier. (UNVERIFIED) tetrameric state An oligomeric state in which 5 or more copies of a protein chain join together to form the working unit. Assemblies this large are usually symmetric rings or shells, so the whole is built by repeating 1 subunit. (UNVERIFIED) higher-order oligomeric state A conformational state of a coronavirus spike protein describing how many of its receptor-binding domains are raised into the open position, ready to engage a host cell. spike opening conformational state A system classification label naming the specific strain or sequence variant of the organism that the simulated molecule comes from, for example a particular variant of the COVID-19 virus. It groups simulations by biological origin. biological strain classification label A system classification label naming a non-standard version of a histone protein present in the structure (for example H2A.Z). Different histone versions can change how DNA is packaged and read. histone variant classification label A nucleic acid analysis tool that characterises RNA and DNA secondary structure and higher-order motifs directly from three-dimensional coordinates. Built on the x3DNA suite, it annotates base pairs, helices, stems, hairpin loops and multi-branch junctions, and produces schematic representations of the result. It is widely used for RNA structural bioinformatics and is integrated into viewers such as Jmol and PyMOL. Dissecting the Spatial Structure of RNA DSSR An information content entity that records which atoms of a molecular system a tool should act on. It may be a stored list of atom indices or an expression that is evaluated to produce that list. It says what to operate on, not what to do with it. (UNVERIFIED) atom selection specification OBSOLETE. An information content entity that records the conditions and factors used when choosing which items to include in a study, such as which ligands to simulate. It states what was considered rather than what was chosen. Retired 2026-08-13, the day after it was minted. It was created to group three children of the retired MOLSIM:000310 plan specification, but only one of the three belonged: MOLSIM:002163 atom selection mask and MOLSIM:002233 atom index group went to MOLSIM:002251 atom selection specification instead, and MOLSIM:001035 validated ligand selection criteria turned out to be an object property that never belonged in a class tree. That left one child, MOLSIM:000143 ligand pairs selection criteria, which is itself the group its twelve criteria belong to. A search of the whole ontology found no other term that could join: every other label mentioning selection or criteria is a convergence threshold, an algorithm or an analysis, all correctly placed elsewhere. MOLSIM:000143 was re-parented to IAO:0000030 information content entity, where it sat before. Never present in a release file. obsolete selection criteria true A state that a molecule or molecular system occupies, characterised by its shape, its surroundings, its position on the energy surface, or its electronic configuration. It describes a condition the system is in, rather than a relationship to another molecule or a choice made when the system was built. (UNVERIFIED) molecular state A molecular termini state in which the N-terminus of a polymer chain is closed off with an acetyl group, so that the end carries no electric charge. Capping this way makes a short peptide behave more like a piece cut from the middle of a longer chain. (UNVERIFIED) ACE acetyl capping group A molecular termini state in which the C-terminus of a polymer chain is closed off with an N-methylamide group, so that the end carries no electric charge. It is the usual partner to an acetyl group at the other end of the chain. (UNVERIFIED) NME N-methylamide capping group A molecular termini state in which the C-terminus of a polymer chain is closed off with a plain amide group, so that the end carries no electric charge. It is a smaller alternative to an N-methylamide cap. (UNVERIFIED) NHE amide capping group A molecular termini state in which the N-terminus of a polymer chain is left uncapped and carries a positive charge, as a free amine group does at the pH of most simulations. This is the usual choice when the chain being simulated is a whole protein rather than a fragment of one. (UNVERIFIED) charged N-terminus A molecular termini state in which the C-terminus of a polymer chain is left uncapped and carries a negative charge, as a free carboxylic acid group does at the pH of most simulations. This is the usual choice when the chain being simulated is a whole protein rather than a fragment of one. (UNVERIFIED) charged C-terminus A non-bonded interaction parameter that specifies the distance beyond which van der Waals interactions are no longer calculated. Several programs set this separately from the electrostatic cut-off so that the two can take different values. (UNVERIFIED) AMBER 26 Reference Manual, p. 499, entry vdw_cutoff. States that es_cutoff and vdw_cutoff control the cutoffs for vdw and electrostatic direct force interactions separately, that cut must not be given as well, and that vdw_cutoff must be greater than or equal to es_cutoff. GROMACS 2026.3 Reference Manual, p. 53, entry rvdw, which gives the distance for the LJ or Buckingham cut-off. rvdw vdw_cutoff van der Waals cutoff distance rvdw vdw_cutoff A publication that surveys and summarises existing published work on a topic rather than reporting new primary results. review article A publication that forms one titled part of an edited book, usually written by different authors from the other parts. book chapter A document presenting the research carried out for an academic degree, submitted to and examined by the awarding institution. Simulation results are often reported here before they appear in any journal. thesis A publication that is one of the collected contributions to a scientific meeting. conference proceeding An information content entity that records one property of a data source, such as how its data was produced or who may see it. These are deliberately kept apart from the data source hierarchy itself, which sorts sources only by what kind of resource they are, so that a single source can carry several descriptors at once. data source descriptor A data source descriptor that records how the data held by a source was produced. data origin A data origin recording that the data was obtained by measurement in a laboratory rather than by calculation. experimental data origin A data origin recording that the data was produced by calculation or simulation rather than by laboratory measurement. computational data origin A data source descriptor that records who may obtain the data held by a source. data access A data access descriptor recording that the data may be obtained by anyone without permission or licence. public access A data access descriptor recording that the data may be obtained only under some condition, such as an institutional membership, a licence, an embargo period, or a direct request to its holder. restricted access An Extended PDB ID is the unique twelve-character code composed of a 'pdb_' header followed by 8 alphanumeric characters. Every Extended PDB ID is assigned to a experimentally determined biomolecular structure deposited in the Protein Data Bank (PDB). This identifier serves to reference and retrieve PDB data. This format is the successor of the legacy four-character PDB ID. Extended PDB ID An InChIKey is a fixed-length, 27-character alphanumeric hash derived from the InChI (International Chemical Identifier) code of a chemical structure, used to uniquely and compactly identify chemical compounds in databases and web searches. InChiKey code InChIKey code A multiscale simulation method that combines different levels of physical detail within a single simulation system. It can involve an explicit atomistic region (high resolution), a coarse-grained region (low resolution) and a hybrid transition region between them. AdResS adaptive resolution scheme An SI unit of time equal to 10^-15 seconds or one quadrillionth of a second. (UNVERIFIED) fs femtosecond A unit of pressure defined as 101,325 pascals (Pa) or 1.01325 bar, approximately equal to the average atmospheric pressure at sea level on Earth. (UNVERIFIED) atm atmosphere A unit of velocity, representing one angstrom (10^-10 meters) of distance traveled per nanosecond (10^-9 seconds) of time. It is commonly used in molecular dynamics simulations and studies of atomic-scale processes. (UNVERIFIED) Å/ns|angstrom/nanosecond angstrom per nanosecond A unit of energy per amount of substance, defined as one kilocalorie of energy (1000 thermochemical gram calories) per one mole of substance. It is commonly used in chemistry and biology for thermodynamic quantities. (UNVERIFIED) kcal/mol|kilocalorie/mole kilocalorie per mole A unit of energy per amount of substance per area, commonly used in molecular dynamics simulations for expressing force constants and restraint weights. (UNVERIFIED) kcal/mol/Å^2|kilocalorie/mole/angstrom^2 kilocalorie per mole per square angstrom A unit of electric charge defined as the charge carried by a single proton. It is a fundamental physical constant, with a value of exactly 1.602176634 × 10^-19 coulombs. (UNVERIFIED) e elementary charge A unit that measures a fluid's resistance to flow under the influence of gravity. It is the ratio of the fluid's dynamic viscosity to its density. (UNVERIFIED) kinematic viscosity unit A centimeter-gram-second (CGS) unit of kinematic viscosity, defined as the kinematic viscosity of a fluid that passes one square centimeter in one second. It is equivalent to the unit Stokes (St). (UNVERIFIED) cm^2/s|stokes|St square centimeter per second A unit that measures the separation of positive and negative electrical charges within a system, indicating the system's overall polarity. (UNVERIFIED) electric dipole moment unit A centimeter-gram-second (CGS) unit of electric dipole moment, defined as 10^-18 statcoulomb-centimeters. One Debye is approximately equal to 3.33564 × 10^-30 coulomb-meters. (UNVERIFIED) D debye An algorithm that finds the rotation bringing 2 matched sets of points into the closest possible agreement, by decomposing the matrix of their cross-correlations. It gives the rotation only, so the 2 sets must already have been centred on their averages. It is the standard way structures are superposed before any difference between them is measured. (UNVERIFIED) Kabsch RMSD algorithm An algorithm that assigns secondary structure to each residue of a protein from its atomic coordinates. It combines hydrogen bond energies with backbone angle preferences, where the older DSSP method uses hydrogen bonds alone. The extra criterion makes its assignments agree more closely with those a crystallographer would make by eye. (UNVERIFIED) STRIDE algorithm A statistical algorithm that describes a distribution as a weighted sum of several bell-shaped components, fitting the position, width and weight of each. It suits data made of several overlapping groups, where one component would fit none of them. Each component can then be treated as a conformational state. (UNVERIFIED) Gaussian mixture model algorithm A statistical algorithm that fits a chosen family of probability distributions, such as the normal distribution, to data by estimating its parameters, either by maximum likelihood or by Bayesian inference with a prior. Assuming a family lets a whole distribution be summarised by 2 or 3 numbers. The assumption is also the weakness: if the data do not follow that family, the fitted distribution is wrong however precise the fit appears. (UNVERIFIED) parametric distribution fitting algorithm A statistical algorithm that estimates a distribution by fitting smooth polynomial pieces joined at chosen points, rather than assuming a named family. It can follow a shape that no standard distribution matches. Its accuracy depends on how many joining points are used, which must be chosen to avoid following noise. (UNVERIFIED) spline-based estimation algorithm An algorithm that computes the surface area of a molecule reachable by solvent, by placing test points on a sphere around each atom, enlarged by the radius of a solvent probe, and counting those not buried by neighbouring atoms. The test points approximate rolling a solvent-sized sphere over the molecule. Its accuracy depends mainly on how many test points are used. (UNVERIFIED) Shrake-Rupley algorithm A similarity score that measures how well the chemical features of 2 aligned molecules, such as hydrogen bond donors, acceptors and charged groups, overlap in space, as a Tanimoto ratio from 0 to 1. It complements a shape comparison, which cannot tell a hydrogen bond donor from an acceptor of the same size. "Color" here refers to feature type, not to any visual property. (UNVERIFIED) color Tanimoto score A similarity score that measures how much the volumes of 2 aligned molecules overlap, as the shared volume divided by the total volume occupied by at least 1 of them. It ranges from 0 for no overlap to 1 for identical shape. It considers only shape, so 2 chemically unlike molecules that fill the same space score highly. (UNVERIFIED) shape Tanimoto score A similarity score adding the shape score and the chemical feature score, giving a value between 0 and 2. Adding them weights shape and chemistry equally, which suits searching for a molecule that both fits a site and makes the same contacts. Reporting the 2 parts separately alongside it is usual, since the sum hides which one is high. (UNVERIFIED) Tanimoto combo score A molecular trajectory analysis format is a specialized file structure designed to store processed data derived from the analysis of a molecular dynamics simulation. These formats typically contain derived quantities, such as secondary structure timelines, pairwise distance matrices, or principal component vectors, rather than raw atomic coordinates. They allow for the efficient storage and rapid retrieval of complex analysis results without the need to re-process the massive original trajectory files. molecular trajectory analysis format A data format that stores spatial or statistical data describing how solvent molecules are distributed around a solute and what properties they carry there. Files in these formats usually hold a 3-dimensional grid of values such as local solvent density, charge distribution, or interaction energy. They are used to visualize hydration shells and to calculate thermodynamic properties of solvation. (UNVERIFIED) solvent analysis format A data format that stores the results of an electronic structure calculation, such as molecular orbitals, electron densities, basis sets or electrostatic potentials. Files in these formats are how quantum chemistry programs hand their results to visualization software and to other analysis tools. The group covers both program-specific formats and interchange formats. (UNVERIFIED) quantum chemistry format A data source that holds data, observations or protocols recorded directly in a digital laboratory notebook system used by a research group. Unlike a public database or a formal publication, it contains raw internal records of day-to-day work and of the exact conditions of an experiment. Citing it says where the data came from when that data sits inside a system owned by one institution and has not yet been publicly deposited or peer reviewed. electronic lab journal source electronic laboratory notebook source A computing system made of networked computers that are housed and administered within a single institution or research group and used as one pooled resource. It differs from a national facility in being controlled locally, so access, hardware and software are decided by the group that runs it. (UNVERIFIED) local computing cluster system An energy analysis that estimates the free energy of a single molecular conformation as the sum of its internal potential energy and its solvation free energy. The internal (conformational) entropy of the solute is left out or estimated separately. It is the basis of end-point methods such as MM-PBSA and MM-GBSA, which use it to estimate binding free energies. (UNVERIFIED) effective energy analysis A computational method that predicts the structure of a complex between a protein and a short peptide. It is harder than docking a small rigid molecule, because a peptide has many rotatable bonds and usually has no fixed shape of its own until it binds. Methods for it therefore search over the peptide's conformation as well as its position. (UNVERIFIED) protein-peptide docking analysis Membrane placement analysis is a computational method used to determine the optimal position and orientation of a protein structure within a lipid bilayer. It calculates the most energetically favorable alignment by assessing the hydrophobicity of the protein surface and matching it to the hydrophobic core and polar interfaces of the membrane. This analysis is a critical prerequisite for setting up stable membrane protein simulations and for understanding how proteins anchor themselves in the cell membrane. membrane placement analysis A validation method in which the LEaP program of AmberTools checks an assembled molecular system for problems before its topology and coordinate files are saved. It reports long or short bonds, a total charge that is not a whole number, missing force field atom types, and non-bonded atoms closer than 1.5 Å. Fixing these first avoids running a simulation on a faulty topology. (UNVERIFIED) leap analysis A sequence_feature with an extent greater than zero. A nucleotide region is composed of bases and a polypeptide region is composed of amino acids. sequence sequence SO:0000001 region A sequence_feature with an extent greater than zero. A nucleotide region is composed of bases and a polypeptide region is composed of amino acids. SO:ke Any extent of continuous biological sequence. INSDC_feature:misc_feature INSDC_note:other INSDC_note:sequence_feature located_sequence_feature sequence feature sequence located sequence feature SO:0000110 sequence_feature Any extent of continuous biological sequence. LAMHDI:mb SO:ke A methylated deoxy-cytosine. methylated C methylated cytosine methylated cytosine base methylated cytosine residue methylated_C sequence SO:0000114 methylated_cytosine A methylated deoxy-cytosine. SO:ke A modified base in which adenine has been methylated. methylated A methylated adenine methylated adenine base methylated adenine residue methylated_A sequence SO:0000161 methylated_adenine A modified base in which adenine has been methylated. SO:ke A modified nucleotide, i.e. a nucleotide other than A, T, C. G. INSDC_feature:modified_base modified base site sequence SO:0000305 Modified base:<modified_base>. modified_DNA_base A modified nucleotide, i.e. a nucleotide other than A, T, C. G. http://www.insdc.org/files/feature_table.html A nucleotide modified by methylation. methylated base feature sequence SO:0000306 methylated_DNA_base_feature A nucleotide modified by methylation. SO:ke A structurally or functionally defined protein region. In proteins with multiple domains, the combination of the domains determines the function of the protein. A region which has been shown to recur throughout evolution. BS:00012 BS:00134 SO:0001069 domain structural domain polypeptide domain polypeptide_structural_domain sequence SO:0000417 Range. Old definition from before biosapiens: A region of a single polypeptide chain that folds into an independent unit and exhibits biological activity. A polypeptide chain may have multiple domains. polypeptide_domain A structurally or functionally defined protein region. In proteins with multiple domains, the combination of the domains determines the function of the protein. A region which has been shown to recur throughout evolution. EBIBS:GAR domain uniprot:feature_type structural domain polypeptide_structural_domain A region of a gene. gene member region sequence SO:0000831 A manufactured term used to allow the parts of a gene to have an is_a path to the root. gene_member_region A region of a gene. SO:ke Biological sequence region that can be assigned to a specific subsequence of a polypeptide. BS:00124 BS:00331 region site sequence positional positional polypeptide feature region or site annotation SO:0000839 Added to allow the polypeptide regions to have is_a paths back to the root. polypeptide_region Biological sequence region that can be assigned to a specific subsequence of a polypeptide. SO:GAR SO:ke region uniprot:feature_type site uniprot:feature_type Region of polypeptide with a given structural property. BS:00337 polypeptide structural region sequence structural_region SO:0001070 Range. polypeptide_structural_region Region of polypeptide with a given structural property. EBIBS:GAR SO:cb A region of peptide with secondary structure has hydrogen bonding along the peptide chain that causes a defined conformation of the chain. BS:00003 http://en.wikipedia.org/wiki/Secondary_structure polypeptide secondary structure sequence 2nary structure secondary structure secondary structure region secondary_structure SO:0001078 Biosapien term was secondary_structure. polypeptide_secondary_structure A region of peptide with secondary structure has hydrogen bonding along the peptide chain that causes a defined conformation of the chain. EBIBS:GAR http://en.wikipedia.org/wiki/Secondary_structure wiki 2nary structure secondary structure secondary structure region secondary_structure A helix is a secondary_structure conformation where the peptide backbone forms a coil. BS:00152 sequence helix SO:0001114 Range. peptide_helix A helix is a secondary_structure conformation where the peptide backbone forms a coil. EBIBS:GAR helix The helix has 3.6 residues per turn which corresponds to a translation of 1.5 angstroms (= 0.15 nm) along the helical axis. Every backbone N-H group donates a hydrogen bond to the backbone C=O group of the amino acid four residues earlier. BS:00040 http://en.wikipedia.org/wiki/Alpha_helix sequence a-helix helix SO:0001117 Range. alpha_helix The helix has 3.6 residues per turn which corresponds to a translation of 1.5 angstroms (= 0.15 nm) along the helical axis. Every backbone N-H group donates a hydrogen bond to the backbone C=O group of the amino acid four residues earlier. EBIBS:GAR http://en.wikipedia.org/wiki/Alpha_helix wiki a-helix helix uniprot:feature_type The pi helix has 4.1 residues per turn and a translation of 1.15 (=0.115 nm) along the helical axis. The N-H group of an amino acid forms a hydrogen bond with the C=O group of the amino acid five residues earlier. BS:00153 http://en.wikipedia.org/wiki/Pi_helix pi helix sequence SO:0001118 Range. pi_helix The pi helix has 4.1 residues per turn and a translation of 1.15 (=0.115 nm) along the helical axis. The N-H group of an amino acid forms a hydrogen bond with the C=O group of the amino acid five residues earlier. EBIBS:GAR http://en.wikipedia.org/wiki/Pi_helix wiki The 3-10 helix has 3 residues per turn with a translation of 2.0 angstroms (=0.2 nm) along the helical axis. The N-H group of an amino acid forms a hydrogen bond with the C=O group of the amino acid three residues earlier. BS:0000340 http://en.wikipedia.org/wiki/310_helix 3(10) helix 3-10 helix 310 helix three ten helix sequence SO:0001119 Range. three_ten_helix The 3-10 helix has 3 residues per turn with a translation of 2.0 angstroms (=0.2 nm) along the helical axis. The N-H group of an amino acid forms a hydrogen bond with the C=O group of the amino acid three residues earlier. EBIBS:GAR http://en.wikipedia.org/wiki/310_helix wiki A base is a sequence feature that corresponds to a single unit of a nucleotide polymer. http://en.wikipedia.org/wiki/Nucleobase sequence SO:0001236 base A base is a sequence feature that corresponds to a single unit of a nucleotide polymer. SO:ke http://en.wikipedia.org/wiki/Nucleobase wiki A region defined by its disposition to be involved in a biological process. INSDC_misc_feature INSDC_note:biological_region biological region sequence SO:0001411 biological_region A region defined by its disposition to be involved in a biological process. SO:cb A biological DNA region implicated in epigenomic changes caused by mechanisms other than changes in the underlying DNA sequence. This includes, nucleosomal histone post-translational modifications, nucleosome depletion to render DNA accessible and post-replicational base modifications such as cytosine modification. kareneilbeck 2010-03-27T12:02:29Z sequence epigenetically modified region SO:0001720 Moved from is_a biological_region (SO:0001411) to is_a regulatory_region (SO:0005836) on 11 Feb 2021. GREEKC members pointed out that this would be a more appropriate location. See GitHub Issue #530. 11 Feb 2021 updated definition along with addition of epigenomically_modified_region (SO:0002332). Epigenetically modified region is now not inherited while epigenomically modified region is not annotated as inherited. See GitHub Issue #532 and issue #534. epigenetically_modified_region A biological DNA region implicated in epigenomic changes caused by mechanisms other than changes in the underlying DNA sequence. This includes, nucleosomal histone post-translational modifications, nucleosome depletion to render DNA accessible and post-replicational base modifications such as cytosine modification. SO:ke http://en.wikipedia.org/wiki/Epigenetics A cytosine methylated at the 5 carbon. kareneilbeck 2012-10-17T12:46:10Z http://www.insdc.org/files/feature_table.html#7.4.2 http:http://www.pacificbiosciences.com/pdf/WP_Detecting_DNA_Base_Modifications_Using_SMRT_Sequencing.pdf 5 methylcytosine 5-mC m-5C m5c sequence SO:0001918 5_methylcytosine A cytosine methylated at the 5 carbon. SO:rtapella An adenine methylated at the 6 nitrogen. kareneilbeck 2012-10-17T12:54:23Z http:http://www.pacificbiosciences.com/pdf/WP_Detecting_DNA_Base_Modifications_Using_SMRT_Sequencing.pdf 6-mA 6-methyladenine 6mA N6-methyladenine m-6A m6a sequence SO:0001920 N6_methyladenine An adenine methylated at the 6 nitrogen. SO:rtapella A modified DNA cytosine base feature, modified by a hydroxymethyl group at the 5 carbon. kareneilbeck 2013-05-17T05:05:31Z http:http://www.pacificbiosciences.com/pdf/WP_Detecting_DNA_Base_Modifications_Using_SMRT_Sequencing.pdf 5-hmC 5-hydroxymethylcytosine sequence SO:0001960 5_hydroxymethylcytosine A modified DNA cytosine base feature, modified by a hydroxymethyl group at the 5 carbon. SO:ke A modified DNA cytosine base feature, modified by a formyl group at the 5 carbon. kareneilbeck 2013-05-17T05:06:13Z http:http://www.pacificbiosciences.com/pdf/WP_Detecting_DNA_Base_Modifications_Using_SMRT_Sequencing.pdf 5-fC 5-formylcytosine sequence SO:0001961 5_formylcytosine A modified DNA cytosine base feature, modified by a formyl group at the 5 carbon. SO:ke A modified adenine DNA base feature. kareneilbeck 2013-05-20T01:22:30Z sequence SO:0001962 modified_adenine A modified adenine DNA base feature. SO:ke A modified cytosine DNA base feature. kareneilbeck 2013-05-20T01:23:47Z sequence SO:0001963 modified_cytosine A modified cytosine DNA base feature. SO:ke A modified DNA cytosine base feature, modified by a carboxy group at the 5 carbon. kareneilbeck 2013-05-20T01:30:01Z http:http://www.pacificbiosciences.com/pdf/WP_Detecting_DNA_Base_Modifications_Using_SMRT_Sequencing.pdf 5-caC 5-carboxycytosine sequence SO:0001966 5_carboxylcytosine A modified DNA cytosine base feature, modified by a carboxy group at the 5 carbon. SO:ke A region of sequence that is involved in the control of a biological process. INSDC_feature:regulatory http://en.wikipedia.org/wiki/Regulatory_region INSDC_qualifier:other regulatory region sequence SO:0005836 regulatory_region A region of sequence that is involved in the control of a biological process. SO:ke http://en.wikipedia.org/wiki/Regulatory_region wiki A region of polypeptide chain with high conformational flexibility. IDR intrinsically disordered region intrinsically unstructured polypeptide region sequence disordered region SO:0100003 intrinsically_unstructured_polypeptide_region A region of polypeptide chain with high conformational flexibility. EBIBS:GAR IDR intrinsically disordered region disordered region Irregular, unstructured regions of a protein's backbone, as distinct from the regular region (namely alpha helix and beta strand - characterised by specific patterns of main-chain hydrogen bonds). peptide coil sequence coil random coil SO:0100012 peptide_coil Irregular, unstructured regions of a protein's backbone, as distinct from the regular region (namely alpha helix and beta strand - characterised by specific patterns of main-chain hydrogen bonds). EBIBS:GAR coil random coil A subsection of sequence with biological interest that is conserved in different proteins. They may or may not have functional or structural significance within the proteins in which they are found. polypeptide conserved region sequence SO:0100021 polypeptide_conserved_region A subsection of sequence with biological interest that is conserved in different proteins. They may or may not have functional or structural significance within the proteins in which they are found. EBIBS:GAR http://stats.stackexchange.com/questions/50623/r-calculating-mean-and-standard-error-of-mean-for-factors-with-lm-vs-direct The standard error of the mean (SEM) is data item denoting the standard deviation of the sample-mean's estimate of a population mean. It is calculated by dividing the sample standard deviation (i.e., the sample-based estimate of the standard deviation of the population) by the square root of n , the size (number of observations) of the sample. A measure of dispersion applied to means across hypothetical repeated random samples. Alejandra Gonzalez-Beltran Brian S. Alper, Harold Lehmann, Muhammad Afzal, Xing Song, Kenneth Wilkins, Joanne Dehnbostel Orlaith Burke Philippe Rocca-Serra SEM adapted from wikipedia (https://en.wikipedia.org/wiki/Standard_error) SEM scipy.stats.sem(a, axis=0, ddof=1) http://docs.scipy.org/doc/scipy/reference/generated/scipy.stats.sem.html#scipy.stats.sem source: https://github.com/scipy/scipy/blob/v0.15.1/scipy/stats/stats.py#L1928 standard error of the mean https://fevir.net/resources/CodeSystem/27270#STATO:0000037 The correlation coefficient of two variables in a data sample is their covariance divided by the product of their individual standard deviations. It is a normalized measurement of how the two are linearly related. Alejandra Gonzalez-Beltran Orlaith Burke Philippe Rocca-Serra STATO r r statistic correlation coefficient Spearman's rank correlation coefficient is a correlation coefficient which is a nonparametric measure of statistical dependence between two ranked variables. It assesses how well the relationship between two variables can be described using a monotonic function. If there are no repeated data values, a perfect Spearman correlation of +1 or −1 occurs when each of the variables is a perfect monotone function of the other. Spearman's coefficient may be used when the conditions for computing Pearson's correlation are not met (e.g linearity, normality of the 2 continuous variables) but may require a ranking transformation of the variables Alejandra Gonzalez-Beltran Orlaith Burke Philippe Rocca-Serra Spearman's rho http://en.wikipedia.org/wiki/Spearman%27s_rank_correlation_coefficient cor(x, y = NULL, use = "everything",method = c("spearman")) scipy.stats.spearmanr(a, b=None, axis=0) http://docs.scipy.org/doc/scipy/reference/generated/scipy.stats.spearmanr.html#scipy.stats.spearmanr source: https://github.com/scipy/scipy/blob/v0.15.1/scipy/stats/stats.py#L2643 Spearman's rank correlation coefficient The standard deviation of a random variable, statistical population, data set, or probability distribution is a measure of variation which correspond to the average distance from the mean of the data set to any given point of that dataset. It also corresponds to the square root of its variance. Alejandra Gonzalez-Beltran Orlaith Burke Philippe Rocca-Serra σ http://en.wikipedia.org/wiki/Standard_deviation sd(x, na.rm = FALSE) http://stat.ethz.ch/R-manual/R-patched/library/stats/html/sd.html standard deviation The Pearson's correlation coefficient is a correlation coefficient which evaluates two continuous variables for association strength in a data sample. It assumes that both variables are normally distributed and linearity exists. The coefficient is calculated by dividing their covariance with the product of their individual standard deviations. It is a normalized measurement of how the two are linearly related. Alejandra Gonzalez-Beltran Orlaith Burke Philippe Rocca-Serra Pearson product-moment correlation coefficient Pearson's r r statistics STATO, adapted from http://www.r-tutor.com/elementary-statistics/numerical-measures/correlation-coefficient and from: http://stamash.org/pearsons-correlation-coefficient/ http://stamash.org/kendalls-tau-correlation/ cor(x, y = NULL, use = "everything",method = c("pearson")) http://stat.ethz.ch/R-manual/R-patched/library/stats/html/cor.html scipy.stats.pearsonr(x, y) http://docs.scipy.org/doc/scipy/reference/generated/scipy.stats.pearsonr.html#scipy.stats.pearsonr source: https://github.com/scipy/scipy/blob/v0.15.1/scipy/stats/stats.py#L2427 Pearson's correlation coefficient The coefficient of determination is a data item measuring the proportion of the variance in the dependent variable that is predictable from the independent variable(s). In the case of a linear regression mode, the coefficient of determination r2 is the quotient of the variances of the fitted values and observed values of the dependent variable. r2 > eruption.lm = lm(eruptions ~ waiting, data=faithful) > summary(eruption.lm)$r.squared from: http://www.r-tutor.com/elementary-statistics/simple-linear-regression/coefficient-determination coefficient of determination A hypothesis testing measure that represents the probability of obtaining a result at least as far from the value actually obtained as the value expected, assuming the null hypothesis is true. Within the frequentist framework, a p-value is typically a p-value for two-sided test and the criterion for rejecting the null hypothesis is typically expressed as a p-value that is less than an alpha setting divided by 2. In some cases, the p-value is a p-value for one-sided test, and the criterion for rejecting the null hypothesis is typically expressed as a p-value that is less than an alpha setting. A p-value is preferably coded more precisely as a p-value for two-sided test or a p-value for one-sided test rather than using the code for p-value without specification. The code for p-value (without specification) may be used in contexts where the p-value reported is ambiguous. Within the Bayesian framework, use the posterior predictive p-value. P value P-value PValue p value SEVCO group p-value A unit of measurement is a standardized quantity of a physical quality. unit A unit of measurement is a standardized quantity of a physical quality. Wikipedia:Wikipedia A unit which is a standard measure of the distance between two points. length unit A unit which is a standard measure of the distance between two points. Wikipedia:Wikipedia A unit which is a standard measure of the amount of matter/energy of a physical object. mass unit A unit which is a standard measure of the amount of matter/energy of a physical object. Wikipedia:Wikipedia A unit which is a standard measure of the dimension in which events occur in sequence. time derived unit time unit A unit which is a standard measure of the dimension in which events occur in sequence. Wikipedia:Wikipedia A unit which is a standard measure of the average kinetic energy of the particles in a sample of matter. temperature derived unit temperature unit A unit which is a standard measure of the average kinetic energy of the particles in a sample of matter. Wikipedia:Wikipedia A time unit which is equal to the duration of 9 192 631 770 periods of the radiation corresponding to the transition between the two hyperfine levels of the ground state of the caesium 133 atom. s second A time unit which is equal to the duration of 9 192 631 770 periods of the radiation corresponding to the transition between the two hyperfine levels of the ground state of the caesium 133 atom. BIPM:BIPM NIST:NIST A thermodynamic temperature unit which is equal to the fraction 1/273.16 of the thermodynamic temperature of the triple point of water. K kelvin A thermodynamic temperature unit which is equal to the fraction 1/273.16 of the thermodynamic temperature of the triple point of water. BIPM:BIPM NIST:NIST A length unit which is equal to one thousandth of one millionth of a meter or 10^[-9] m. nanometre nm nanometer A length unit which is equal to one thousandth of one millionth of a meter or 10^[-9] m. NIST:NIST A length unit which is equal to 10 [-10] m. Å angstrom A length unit which is equal to 10 [-10] m. NIST:NIST A time unit which is equal to one thousandth of a second or 10^[-3] s. ms millisecond A time unit which is equal to one thousandth of a second or 10^[-3] s. NIST:NIST A time unit which is equal to one millionth of a second or 10^[-6] s. us microsecond A time unit which is equal to one millionth of a second or 10^[-6] s. NIST:NIST A time unit which is equal to 10^[-12] s. ps picosecond A time unit which is equal to 10^[-12] s. NIST:NIST A unit which is one of a particular measure to which all measures of that type can be related. base unit A unit which is one of a particular measure to which all measures of that type can be related. NIST:NIST A unit which is a standard measure of the amount of a 2-dimensional flat surface. area unit A unit which is a standard measure of the amount of a 2-dimensional flat surface. UOC:GVG A unit which represents a standard measurement of how much of a given substance there is mixed with another substance. concentration unit A unit which represents a standard measurement of how much of a given substance there is mixed with another substance. UOC:GVG A density unit which is a standard measure of the mass of a substance in a given volume. mass per unit volume mass density unit A density unit which is a standard measure of the mass of a substance in a given volume. UOC:GVG A unit which is a standard measure of the rate of movement. Speed is measured in the same physical units of measurement as velocity, but does not contain the element of direction that velocity has. Speed is thus the magnitude component of velocity. speed/velocity unit A unit which is a standard measure of the rate of movement. Speed is measured in the same physical units of measurement as velocity, but does not contain the element of direction that velocity has. Speed is thus the magnitude component of velocity. Wikipedia:Wikipedia A concentration unit which is a standard measure of the number of moles of a given substance per liter of solution. unit of molarity A concentration unit which is a standard measure of the number of moles of a given substance per liter of solution. UOC:GVG A unit of concentration which expresses a concentration of 1 mole of solute per liter of solution (mol/L). M molar A unit of concentration which expresses a concentration of 1 mole of solute per liter of solution (mol/L). UOC:GVG A unit of molarity which is equal to one thousandth of a molar or 10^[-3] M. mM millimolar A unit of molarity which is equal to one thousandth of a molar or 10^[-3] M. UOC:GVG A speed/velocity unit which is equal to the speed of an object traveling 1 meter distance in one second. m/s metre per second meter per second A speed/velocity unit which is equal to the speed of an object traveling 1 meter distance in one second. NIST:NIST A unit which is a standard measure of the amount of space occupied by any substance, whether solid, liquid, or gas. volume unit A unit which is a standard measure of the amount of space occupied by any substance, whether solid, liquid, or gas. NIST:NIST A unit which is a standard measure of the number of repetitive actions in a particular time. frequency unit A unit which is a standard measure of the number of repetitive actions in a particular time. NIST:NIST A frequency unit which is equal to 1 complete cycle of a recurring phenomenon in 1 second. Hz s^1 hertz A frequency unit which is equal to 1 complete cycle of a recurring phenomenon in 1 second. NIST:NIST A unit which is a standard measure of the force is applied when a mass is accelerated. force unit A unit which is a standard measure of the force is applied when a mass is accelerated. NIST:NIST A unit which is a standard measure of the force applied to a given area. pressure unit A unit which is a standard measure of the force applied to a given area. NIST:NIST A unit which is a standard measure of the work done by a certain force (gravitational, electric, magnetic, force of inertia, etc). energy unit A unit which is a standard measure of the work done by a certain force (gravitational, electric, magnetic, force of inertia, etc). NIST:NIST A unit which is a standard measure of the figure or space formed by the junction of two lines or planes. angle unit A unit which is a standard measure of the figure or space formed by the junction of two lines or planes. Wikipedia:Wikipedia A unit which is a standard measure of the angle formed by two straight lines in the same plane. plane angle unit A unit which is a standard measure of the angle formed by two straight lines in the same plane. Wikipedia:Wikipedia A plane angle unit which is equal to the angle subtended at the center of a circle by an arc equal in length to the radius of the circle, approximately 57 degrees 17 minutes and 44.6 seconds. rad radian A plane angle unit which is equal to the angle subtended at the center of a circle by an arc equal in length to the radius of the circle, approximately 57 degrees 17 minutes and 44.6 seconds. NIST:NIST A time unit which is equal to one thousandth of one millionth of a second or 10^[-9] s. ns nanosecond A time unit which is equal to one thousandth of one millionth of a second or 10^[-9] s. NIST:NIST A unit which is a standard measure of the influence exerted by some mass. density unit A unit which is a standard measure of the influence exerted by some mass. Wikipedia:Wikipedia A plane angle unit which is equal to 1/360 of a full rotation or 1.7453310^[-2] rad. degree A plane angle unit which is equal to 1/360 of a full rotation or 1.7453310^[-2] rad. Wikipedia:Wikipedia A unit which is a standard measure of physical quantity consisting of only a numerical number without any units. dimensionless unit A unit which is a standard measure of physical quantity consisting of only a numerical number without any units. Wikipedia:Wikipedia A dimensionless unit which denotes a simple count of things. count count unit A dimensionless unit which denotes a simple count of things. MGED:MGED A unit which is a standard measure of the quantity of unbalanced electricity in a body (either positive or negative) and construed as an excess or deficiency of electrons. electric charge A unit which is a standard measure of the quantity of unbalanced electricity in a body (either positive or negative) and construed as an excess or deficiency of electrons. WordNet:WordNet An independently to the base SI units defined mass unit which is equal to one twelfth of the mass of an unbound atom of the carbon-12 nuclide, at rest and in its ground state. Da amu u unified atomic mass unit dalton An independently to the base SI units defined mass unit which is equal to one twelfth of the mass of an unbound atom of the carbon-12 nuclide, at rest and in its ground state. Wikipedia:Wikipedia The electric field strength is a unit which is a measure of the potential difference between two points some distance apart. E-field strength electric field strength unit The electric field strength is a unit which is a measure of the potential difference between two points some distance apart. Wikipedia:Electric_field The volt per meter is a unit of electric field strength equal to the a potential difference of 1 volt existing between two points that are 1 meter apart. V/m volt per metre volt per meter The volt per meter is a unit of electric field strength equal to the a potential difference of 1 volt existing between two points that are 1 meter apart. Wikipedia:Electric_field A unit which represents a standard measurement occurrence of a process per unit time. rate unit A unit which represents a standard measurement occurrence of a process per unit time. UOC:GVG An area unit which is equal to an area enclosed by a square with sides each 1 angstrom long. A^[2] square angstrom An area unit which is equal to an area enclosed by a square with sides each 1 angstrom long. UOC:GVG A frequency unit which is equal to one million hertz or 10^[6] V. mH megahertz megaHertz A frequency unit which is equal to one million hertz or 10^[6] V. UOC:GVG A pressure unit which is equal to 100,000 pascal or the pressure or stress on a surface caused by a force of 100,000 newtons spread over a surface of 1 m^[2]. bar A pressure unit which is equal to 100,000 pascal or the pressure or stress on a surface caused by a force of 100,000 newtons spread over a surface of 1 m^[2]. NIST:NIST meter based unit second based unit kelvin based unit angstrom based unit molar based unit meter per second based unit hertz based unit pascal based unit radian based unit degree based unit dalton based unit volt per meter based unit example to be eventually removed The term was initially used in an attempt to structure part of the ontology, but in retrospect failed to do a good job. failed exploratory term Class has all its metadata, but is either not guaranteed to be in its final location in the asserted IS_A hierarchy or refers to another class that is not complete. metadata complete The term was created to ease viewing/sorting terms for development purposes, but will not be included in a release. organizational term Class has undergone final review, is ready for use, and will be included in the next release. Any class lacking "ready_for_release" should be considered likely to change place in hierarchy, have its definition refined, or be obsoleted in the next release. Those classes deemed "ready_for_release" will also derived from a chain of ancestor classes that are also "ready_for_release." ready for release Class is being worked on; however, the metadata (including definition) are not complete or sufficiently clear to the branch editors. metadata incomplete Nothing done yet beyond assigning a unique class ID and proposing a preferred term. uncurated All definitions, placement in the asserted IS_A hierarchy and required minimal metadata are complete. The class is awaiting a final review by someone other than the term editor. pending final vetting The term was created to temporarily stand in for a semantic purpose, but is no longer needed, typically due to another permanent term being defined. placeholder removed The term has been combined with one or more other terms to create a more encompassing (merged) term. terms merged The term has been replaced by a term imported from another ontology. term imported The term has been split into two or more new terms. term split A Formal Theory of Substances, Qualities, and Universals, http://ontology.buffalo.edu/bfo/SQU.pdf universal A defined class is a class that is defined by a set of logically necessary and sufficient conditions but is not a universal defined class A named class expression is a logical expression that is given a name. The name can be used in place of the expression. named class expression Terms with this status should eventually replaced with a term from another ontology. group:OBI to be replaced with external ontology term A term that is metadata complete, has been reviewed, and problems have been identified that require discussion before release. Such a term requires editor note(s) to identify the outstanding issues. group:OBI requires discussion A protein structural database that is the single worldwide archive for experimentally determined 3-dimensional structures of large biological molecules, including proteins, nucleic acids and their complexes. Many simulations of biological molecules begin with a structure taken from it. Its accession codes are the standard way a structure is cited. (UNVERIFIED) Protein Data Bank PDB A protein structural database that is the European member of the worldwide Protein Data Bank partnership, run by EMBL-EBI, holding the same archive and adding its own search and analysis tools. The members of the partnership share 1 set of entries, so a structure deposited through any of them appears at all. What differs is the interface and the services built on top. (UNVERIFIED) Protein Data Bank in Europe PDBe A protein structural database that is the Japanese member, and a founding member, of the worldwide Protein Data Bank partnership, run at the Institute for Protein Research of Osaka University. It processes the structures deposited mainly from Asia and the Middle East, holds the same archive as the other members, and develops its own tools and services. (UNVERIFIED) PDBj A protein structural database that gives a pictorial summary of each 3D structure in the Protein Data Bank, with schematic diagrams of molecular interactions, protein domains, ligand binding sites and other structural features. It helps a reader understand the contents of an entry quickly, without opening the coordinates. (UNVERIFIED) PDBsum The Nucleic Acid Database (NDB) was a specialized, value-added web repository founded in 1991 to collect, archive, and distribute three-dimensional (3D) structural information about DNA, RNA, and their protein/ligand complexes. The legacy NDB resource has been officially retired. It has been replaced by the Nucleic Acid Knowledgebase (NAKB), which preserves all original NDB functionality while expanding coverage to modern structural biology data. Nucleic Acid Database NDB The Electron Microscopy Data Bank (EMDB) is a global public repository for 3D electron microscopy (3DEM) density maps and tomograms of biological specimens, macromolecules, viruses, and subcellular structures. It stores cryogenic-sample electron microscopy (cryoEM) volumes, helical reconstructions, electron tomograms, subtomogram averages, and electron crystallography maps. Electron Microscopy Data Bank EMDB An electron microscopy database storing the raw, 2D micrographs and tilt series that researchers use to build 3D cryo-electron microscopy (cryo-EM) maps and models (which are archived in the EMDB). Electron Microscopy Public Image Archive EMPIAR A nucleic acid sequence database of the U.S. National Institutes of Health, maintained by the National Center for Biotechnology Information (NCBI), that holds an annotated collection of all publicly available DNA sequences. It is the American member of the International Nucleotide Sequence Database Collaboration, together with ENA in Europe and DDBJ in Japan, and the 3 exchange records daily. (UNVERIFIED) GenBank A knowledge-based database of protein sequences together with what is known about each protein's function, structure and modifications, run by the UniProt consortium. It was formed by joining Swiss-Prot, TrEMBL and the PIR Protein Sequence Database, and it has a manually reviewed section, Swiss-Prot, and a much larger automatically annotated section, TrEMBL. (UNVERIFIED) UniProt Knowledgebase UniProtKB A nucleic acid sequence database, run by EMBL-EBI, covering raw sequencing reads, assembled sequences and their annotations from organisms of every kind. It is the European member of the International Nucleotide Sequence Database Collaboration, whose 3 members exchange records daily, so a sequence deposited at any of them can be found at all 3. (UNVERIFIED) European Nucleotide Archive ENA A nucleic acid sequence database that is the Japanese member of the International Nucleotide Sequence Database Collaboration, together with GenBank in the United States and ENA in Europe. The 3 members exchange records daily, so a sequence deposited at any of them can be found at all 3. (UNVERIFIED) DNA Data Bank of Japan DDBJ A protein sequence database that forms the manually annotated and reviewed section of the UniProt Knowledgebase. Its entries combine the protein sequence with functional annotation and cross-references, checked by curators against the literature, which makes it the most reliable part of UniProtKB. (UNVERIFIED) UniProtKB/Swiss-Prot A protein sequence database that forms the automatically annotated, unreviewed section of the UniProt Knowledgebase. Most of its entries are translations of the coding sequences submitted to the nucleotide sequence databases, and together they give broad coverage of known proteins. (UNVERIFIED) UniProtKB/TrEMBL A protein sequence database, run by NCBI, that provides a curated, non-redundant set of reference protein sequences, often linked to the matching nucleotide sequences and genome annotations. Its records serve as stable reference points for identifying, annotating and comparing proteins. (UNVERIFIED) NCBI RefSeq Protein A protein sequence database whose Protein Sequence Database was one of the first curated collections of protein sequences, growing out of an atlas of protein sequences begun in the 1960s. That collection was merged into the UniProt Knowledgebase, and the resource itself remains a member of the UniProt consortium. (UNVERIFIED) Protein Information Resource PIR A nucleic acid sequence database, run by NCBI, that provides a non-redundant, annotated set of reference sequences for chromosomes, genomic regions, transcripts and RNAs. Each sequence serves as a stable reference for the organism it comes from, which supports genome annotation. (UNVERIFIED) NCBI RefSeq Nucleotide A chemical structure database within PubChem, run by NCBI, that holds 1 record for each unique, validated small-molecule structure, derived from the substance records that many contributors deposit. Each record gives identifiers, computed properties and links to related information. (UNVERIFIED) PubChem Compound A chemical structure database and ontology covering small molecules that take part in biological processes, as products, reactants or regulators. Being an ontology as well as a dictionary, it records how compounds relate to one another rather than only listing them. (UNVERIFIED) Chemical Entities of Biological Interest ChEBI A chemical structure database of commercially available compounds, prepared in ready-to-dock 3D formats for virtual screening, which is the computer search for molecules that may bind a target. It is provided by the Irwin and Shoichet laboratories at the University of California, San Francisco, and holds hundreds of millions of purchasable compounds with annotated properties. (UNVERIFIED) ZINC20 A drug information database holding detailed data on drugs together with their targets, interactions and metabolism. It combines information about each drug with information about the proteins it acts on, and its data are free for non-commercial research but need a licence for commercial use. (UNVERIFIED) DrugBank A chemical structure database, owned by the Royal Society of Chemistry, that gives free access to millions of chemical structures with their properties and related information, gathered from many sources. It can be searched by chemical name, synonym, trade name, registry number or structure. (UNVERIFIED) ChemSpider A drug information database recording the protein and nucleic acid targets that drugs act on, the drugs directed at each, and the diseases involved. It is organised around the target rather than the compound. It is used to find what has already been tried against a protein before a new study is designed. (UNVERIFIED) Therapeutic Target Database TTD A drug information database, run by the U.S. National Library of Medicine, that publishes the labels of marketed drugs as submitted to the U.S. Food and Drug Administration, including package inserts. The labels describe each drug's ingredients, uses, dosage, warnings and side effects for healthcare providers and the public. (UNVERIFIED) DailyMed A bioactivity database of manually curated data on drug-like small molecules and their measured activity against biological targets, mostly proteins, maintained by EMBL-EBI. It brings together chemical, bioactivity and genomic data to help turn genomic information into new drugs. (UNVERIFIED) ChEMBL A spectroscopic database holding mass spectrometry datasets, chiefly from proteomics experiments. It stores the measurements themselves consenting to check reproducibility. Mass Spectrometry Interactive Virtual Environment MassIVE The Spectral Database for Organic Compounds (SDBS) is a free, searchable online database of experimental molecular spectra for organic molecules, hosted by the National Institute of Advanced Industrial Science and Technology (AIST) in Japan. SDBS contains NMR (1H, 13C), mass spectra, IR, Raman, and ESR spectra for a large number of organic compounds. SDBS An NMR data repository collecting nuclear magnetic resonance measurements on biological macromolecules and metabolites, including assigned chemical shifts, coupling constants and relaxation data. Those quantities are what a simulated structure or trajectory is compared against. Biological Magnetic Resonance Data Bank BMRB An NMR data repository specialising in assigned chemical shifts for complex carbohydrates and their complexes. Carbohydrates are poorly covered by the general NMR archives, because their signals overlap heavily and their assignment is difficult. It exists to hold the assignments that do get made. (UNVERIFIED) Complex Carbohydrate Magnetic Resonance Database ComplexCARP A secondary database, run by EMBL-EBI, that combines the protein family, domain and functional site signatures of several member databases, such as Pfam, PROSITE, SMART, CDD and CATH-Gene3D, into 1 resource. Searching a protein sequence against it shows which families and domains the protein belongs to, which helps predict its function. (UNVERIFIED) InterPro A protein family and domain database in which each family is represented by a multiple sequence alignment and a hidden Markov model, a statistical model of the sequence pattern the family shares. It is used to find conserved domains and families in protein sequences, and it is now provided through InterPro at EMBL-EBI. (UNVERIFIED) Pfam A protein family and domain database for identifying and annotating mobile domains, which are domains found in many different protein contexts, and for analysing the domain architecture of proteins. It is developed at EMBL and holds manually curated models for more than 1,000 domains, many of them involved in signalling, extracellular functions and chromatin regulation. (UNVERIFIED) SMART A protein family and domain database that describes protein domains, families and functional sites by sequence patterns or by profiles, which are scoring tables derived from multiple sequence alignments. It is maintained by the SIB Swiss Institute of Bioinformatics and is used to find known motifs and domains in a protein sequence. (UNVERIFIED) PROSITE A protein family and domain database, maintained by NCBI, holding curated models of protein domains together with models imported from other collections, such as Pfam and SMART. Searching a sequence against it shows which known domains a protein contains, which helps decide what part of a protein to simulate. (UNVERIFIED) Conserved Domain Database CDD A structural classification database that classifies protein domains of known structure by their evolutionary relationships and structural similarities. It is the successor of the original SCOP, developed at the MRC Laboratory of Molecular Biology, and organises its classes as a network in which a domain can be related to more than 1 group, instead of the single strict tree that SCOP used. (UNVERIFIED) SCOP2 A structural classification database that sorts protein domains taken from experimental structures into a hierarchy of 4 levels, from broad shape down to evolutionary relationship. The 4 levels are what its name enumerates. It is used to find structurally similar proteins when sequence similarity is too weak to detect. (UNVERIFIED) Class, Architecture, Topology, Homologous superfamily CATH A structural classification database that groups protein domains by evolutionary relationship rather than by shape, with a focus on distant relationships between proteins whose sequences have become too different to show them. Its levels run from architecture through possible homology (X-groups), homology (H-groups) and topology (T-groups) to families, and it is updated regularly as new structures enter the Protein Data Bank. (UNVERIFIED) Evolutionary Classification of Protein Domains ECOD A pathway database of manually drawn maps of molecular interaction and reaction networks, covering metabolism, genetic and environmental information processing, cellular processes, organismal systems and human diseases. It is part of the Kyoto Encyclopedia of Genes and Genomes (KEGG), developed in Japan by Kanehisa Laboratories. (UNVERIFIED) KEGG Pathway A pathway database of human biological pathways and processes that is free, open-source, curated by experts and peer-reviewed. It describes biology as a hierarchy running from single molecular reactions up to whole pathways, linking the genes, proteins and small molecules involved. (UNVERIFIED) Reactome A pathway database that collects thousands of organism-specific databases, each describing the genome and metabolic pathways of 1 organism, together with MetaCyc, a reference database of pathways from many organisms. It is maintained by SRI International and records genes, enzymes, metabolites and reactions, which supports the comparison of metabolism between organisms. (UNVERIFIED) BioCyc A pathway database of biological pathway diagrams that anyone can contribute to and edit, so the collection is curated by its community of users. It covers many species, and its content is released under the CC0 public domain waiver. (UNVERIFIED) WikiPathways A public, web-accessible database focusing on measured binding affinities of interactions between proteins (considered as drug targets) and small, drug-like molecules. It primarily contains data from enzyme assays, isothermal titration calorimetry, and surface plasmon resonance. BindingDB A free, public repository hosted by the U.S. National Center for Biotechnology Information (NCBI) that archives experimental descriptions of biological activities and test results for small molecules, cell-based screens, and small interfering RNAs (siRNAs). PubChem BioAssay An expert-curated, open-access online portal and searchable database that provides quantitative data on biological targets and the prescription medicines and experimental drugs that act on them. Its entries are written and checked by specialists in each target family rather than gathered automatically. IUPHAR/BPS Guide to Pharmacology IUPHAR/BPS A knowledge-based database that collects gene-disease and variant-disease associations from public databases and the scientific literature, standardises them, and scores each association so that it can be ranked. It is developed by the company MedBioinformatics Solutions, with free access for academic research and paid licences for commercial use. (UNVERIFIED) DisGeNET A knowledge-based database cataloguing human genes and the inherited disorders and traits associated with them. It is organised around the link between a gene and the effects of altering it, rather than around sequences. It can be used to connect a mutation being simulated to its known clinical consequences. (UNVERIFIED) Online Mendelian Inheritance in Man OMIM A repository database for research outputs from all disciplines, such as datasets, software, publications and figures, operated by CERN and created within the European OpenAIRE project. Every upload receives a DOI, a permanent identifier that makes it citable. (UNVERIFIED) Zenodo A repository database, checked by curators, that publishes the data underlying scientific and medical publications so that the data can be found, freely reused and cited. Researchers deposit the datasets that belong to their articles, and the data are released into the public domain under the CC0 dedication. (UNVERIFIED) Dryad Digital Repository A repository database where researchers in any discipline can store and share research outputs such as figures, datasets, images and code. Each item can be given a DOI, a permanent identifier that makes it citable and easier to find. (UNVERIFIED) Figshare A simulation data repository, based in Europe, for storing and sharing biomolecular simulation data. It accepts trajectories together with the metadata describing how they were produced, which is what makes a deposited trajectory reusable by others. (UNVERIFIED) Molecular Dynamics Data Bank MDDBR A simulation data repository for molecular dynamics trajectories of biomolecules, contributed by the biomolecular simulation community in the United Kingdom and hosted by the Physical Sciences Data Infrastructure (PSDI). It records how each simulation was set up, from the starting structure and force field to the run parameters, so that the data can be reproduced and reused. (UNVERIFIED) BioSimDB A simulation data repository of atomistic molecular dynamics trajectories of proteins, chosen to represent the monomeric soluble protein structures in the Protein Data Bank. It was developed by IRB Barcelona and the Barcelona Supercomputing Center, and its trajectories are used to describe how flexible each protein is. (UNVERIFIED) MoDEL database Distinct from the class 'model' (MOLSIM:000006), which it resembles only in spelling: this is a named simulation data repository, that is a general kind of representation. Case-insensitive lexical matching conflates the two, so prefer the synonym 'MoDEL database' when matching. MoDEL A web server tool for sharing, viewing and analysing molecular dynamics simulation trajectories directly in a web browser, without installing software locally. A research group runs it on its own server to make its trajectories available to others, with interactive views and common analyses. (UNVERIFIED) MDsrv A protein structural database for structures of biological molecules determined by integrative or hybrid methods, which combine data from several experimental techniques with computational modelling. It began as a prototype archive and in 2024 was unified with the Protein Data Bank and renamed PDB-IHM, so its structures now receive PDB accession codes. (UNVERIFIED) PDB-Dev A predicted structure repository that gives open access to protein structure predictions made by the AlphaFold system, covering a large share of the protein sequences in UniProt. It is run by EMBL-EBI together with Google DeepMind, and supplies models for proteins that have no experimentally determined structure. (UNVERIFIED) AlphaFold Protein Structure Database A predicted structure repository of 3D protein models built automatically by the SWISS-MODEL homology modelling pipeline, which bases each model on an experimentally determined structure of a related protein. It is updated regularly and provides precomputed models for a large number of sequences in UniProtKB, especially those of model organisms. (UNVERIFIED) SWISS-MODEL Repository A predicted structure repository holding several hundred million structure predictions for metagenomic proteins, which are proteins known only from DNA sequenced directly from environmental samples, made with the ESMFold model developed at Meta AI. ESMFold predicts a structure from a single sequence with a protein language model, without first collecting related sequences, which makes it much faster than earlier methods but on average less accurate. (UNVERIFIED) Evolutionary Scale Modeling Atlas ESMatlas A predicted structure repository for 3D models of biological macromolecules made by computational methods, which the Protein Data Bank does not accept because they are not based on experimental data. Each deposited model receives a stable accession code, which is also a DOI, so that it can be cited, and the archive is developed by the SIB Swiss Institute of Bioinformatics and the Biozentrum of the University of Basel. (UNVERIFIED) ModelArchive A metabolomics repository, maintained at EMBL-EBI, for metabolomics experiments across species and measurement techniques, together with the information derived from them. It archives study descriptions, raw experimental data and the identified and measured metabolites. (UNVERIFIED) MetaboLights A metabolomics repository and analysis platform for mass spectrometry data on natural products, which groups related compounds by the similarity of their spectra. Grouping this way helps suggest the identity of compounds that no reference library contains, from their resemblance to ones that it does. Its reference spectra are contributed and curated by its users. (UNVERIFIED) Global Natural Products Social Molecular Networking GNPS A public repository for metabolomics data and metadata, part of the NIH Common Fund's Metabolomics Program. It serves as a national and international data repository and coordinating center, providing tools for data analysis, visualization, and integration with other omics data. (UNVERIFIED) Metabolomics Workbench A literature database, run by the U.S. National Library of Medicine at the National Institutes of Health, that gives free search access to references and abstracts in the life sciences and biomedicine, drawn mainly from the MEDLINE database. Its records link to the full text where a publisher or PubMed Central makes it available. (UNVERIFIED) PubMed A literature database of abstracts and citations covering peer-reviewed journals, books and conference proceedings, owned by the publisher Elsevier and available by subscription. It provides tools to track, analyse and visualise research output and citations across disciplines. (UNVERIFIED) Scopus A literature database and citation index covering many disciplines, owned by the company Clarivate and available by subscription. It gives access to several databases of scholarly journals, books, conference proceedings and patents, with the citation links between their records, and is used for literature searches and for measuring research impact. (UNVERIFIED) Web of Science A literature database, provided free of charge by Google as a web search engine, that indexes the full text or metadata of scholarly literature across many publishing formats and disciplines. It is not curated: its index is built automatically, so its coverage is broad but its records can contain errors. (UNVERIFIED) Google Scholar The small molecule structural database that serves as the primary repository for experimentally determined three-dimensional crystal structures of small organic and metal-organic molecules. It provides high-quality reference geometries used for validating force fields and for obtaining starting structures in computational studies. (UNVERIFIED) Cambridge Structural Database OBSOLETE. A redundant untyped individual that duplicated the class 'omni-path' (MOLSIM:001949). It should not be used; use that class instead. Obsoleted: redundant untyped individual duplicating the class 'omni-path' (MOLSIM_001949), which is the concept to use instead. obsolete Omni-Path true The file format convention denoted by the string AMBERRESTART in a NetCDF file's global Conventions attribute, designating the file as an AMBER restart or state file. Such a file holds a single configuration with no frame dimension, while a trajectory holds many frames, and because NetCDF is a general container this attribute is the declared way to tell the 2 apart. (UNVERIFIED) AMBERRESTART NetCDF convention The file format convention denoted by the string AMBER in a NetCDF file's global Conventions attribute, designating the file as an AMBER coordinate trajectory. It is the counterpart of the restart convention and distinguishes a file of many frames from a file holding 1 state. The distinction cannot be made from the file extension alone. (UNVERIFIED) AMBER trajectory NetCDF convention The system classification label assigned to a molecular system composed of a protein receptor, a small-molecule ligand and explicit water molecules. Explicit water means that individual water molecules are present in the model, rather than a continuum that only imitates their average effect. (UNVERIFIED) Protein / Ligand in Explicit Water The hydrogen detection strategy that identifies hydrogen atoms by testing whether an atom's atomic number is exactly 1, used when applying or detecting hydrogen mass repartitioning. It is the reliable test, because atomic number is a physical fact recorded in the topology. It cannot be used where the topology omits atomic numbers, which is why a name-based fallback exists. (UNVERIFIED) atomic number HMR detection strategy The hydrogen detection strategy that identifies hydrogen atoms by testing whether an atom's name begins with the letter H, used when applying or detecting hydrogen mass repartitioning. It is a fallback for topologies that do not record atomic numbers. It is not reliable: the element symbols of helium, hafnium and mercury (He, Hf, Hg) also begin with H, and some naming conventions put a digit before the H. (UNVERIFIED) atom name HMR detection strategy The term was added to the ontology on the assumption it was in scope, but it turned out later that it was not. out of scope The term was added to the ontology on the assumption it was a valid domain entity, but it turns out the entity does not exist in reality. domain entity does not exist