Version 0.18.2 ============== BUGFIXES: * Revert universal newlines for CSV parsing * Avoid potential issue with threaded kegg download in python3 Version 0.18.1 ============== BUGFIXES: * Ability to parse the new CSV format for phenotype microarrays * Long kegg URL bugfix Version 0.18.0 ============== FEATURES: * Ability to parse the new CSV format for phenotype microarrays * Compatibility with python 2 and python 3 BUGFIXES: * kegg: fix coloring of compounds (broken since 0.17.18) Version 0.17.18 ============== BUGFIXES: * kegg: try to salvage downloading pathways with many annotations Version 0.17.17 ============== BUGFIXES: * kegg: add a more robust mechanism to ignore kegg generic pathways * kegg: reduce the number of threads for kegg mapping to reduce HTTP 403 errors * enforce the use of python2 Version 0.17.14 ============== BUGFIXES: * kegg: add map01230 to the pathways to ignore Version 0.17.13 ============== UPGRADE URGENCY: LOW, MODERATE if you use the latest version of matplotlib BUGFIXES: * yaml parsing: avoid deprecation warning when parsing * avoid deprecation warning/error when using latest and new-ish matploblib Version 0.17.12 ============== UPGRADE URGENCY: LOW, HIGH if you use the latest version of networkx (2.x) BUGFIXES: * kegg: avoid a crash Version 0.17.10 ============== UPGRADE URGENCY: LOW, HIGH if you use the latest version of matplotlib (2.1.0) BUGFIXES: * kegg: avoid a crash when running a mutant analysis * dphenome plot: avoid crashes due to changes in behavior of numpy/matplotlib Version 0.17.8 ============== UPGRADE URGENCY: HIGH, the KEGG data download step is now broken due to the retirement of the RPAIR database BUGFIXES: * kegg: the data contained in the now discontinued RPAIR database is retrieved from the KEGG reactions DB Version 0.17.6 ============== UPGRADE URGENCY: HIGH if you have the latest version of networkx installed BUGFIXES: * dape start: newer versions of networkx do not cause crashes anymore * Removed hardcoded KEGG API addresses (removes useless warning messages) Version 0.17.4 ============== UPGRADE URGENCY: LOW if you are not experiencing crashes with the "dphenome stats" command BUGFIXES: * Do not smooth/compress PM signals when they are too few (manual experiments) * dphenome stats: avoid crashes for unusual AV distributions Version 0.17.2 ============== UPGRADE URGENCY: LOW if you are not experiencing crashes with the "dape start" command BUGFIXES: * dape start: avoid crashes when recent networkx versions are used * dape/dgenome/dphenome: use "/usr/bin/env python" instead of "/usr/bin/python" Version 0.17.0 ============== UPGRADE URGENCY: LOW if you are not interested in the new features Please note that python 3 is still not supported FEATURES: * dphenome rings: use a parameter of choice instead of the activity * the new -r option can be used * dphenome trim: the time at which perform the trim can be set by the user * dgenome export: organisms are separated by tab (pangenome_category.tsv) BUGFIXES: * dgenome start: more information in the logs Version 0.16.4 ============== UPGRADE URGENCY: LOW if you are not using PM experiments with a low number of time points FEATURES: * dphenome start: no smoothing/compressing when not enough time points * when the number of time points is below 33 * "manual" phenomic experiments may have a very low number of time points BUGFIXES: * Full python3 compatibility * dgenome start still had some python2 bits * dape/dgenome/dphenome start: option "-p" renamed to avoid conflicts * dape start -p --> dape start -s * dgenome start -p PREFIX --> dgenome start -x PREFIX * dphenome start -p --> dphenome start -s * dphenome stats: do not exit if the AV has not been computed yet * The command may still output useful data Version 0.16.2 ============== UPGRADE URGENCY: HIGH if you use the "dape map" command or have the latest matplolib library, LOW otherwise BUGFIXES: * dape map: could not get any colored pathway map * Installation: NetworkX was missing in the setup file * dphenome rings: * Memory-leak when many organisms are present * Different proportions when orgs > 10 * Fix crashes w/ latest matplotlib version Version 0.16.0 ============== UPGRADE URGENCY: MODERATE, one corner-case crash on dphenome plot has been fixed FEATURES: * dphenome plot: it is now possible to plot single plates/wells with one command * dphenome plot PM01 --> plots only plate PM01 (w/ heatmap) * dphenome plot PM01 A01 --> plots only well PM01 A01 BUGFIXES: * dphenome plot: in some cases the creation of the legends caused a crash of the command Version 0.15.0 ============== UPGRADE URGENCY: MODERATE, only feature improvements IMPORTANT NOTE: this version introduces a backward incompatible change in dphenome Users who want to analyze an old DuctApe project files should rebuilt it from scratch (or using the various "export" and "import" commands) FEATURES: * dphenome add: new parsing logic for plate IDs * Before: PM03B --> the final letter was kept, causing confusion and failures * After: PM03 --> more consistent and durable * This is a backward incompatible change * dphenome stats: output the tables to file (not just to the terminal) * dgenome stats: output the tables to file (not just to the terminal) BUGFIXES: * dphenome add: yml files with "strange" plate IDs caused crashes * Print functions compatible with python 3 Version 0.14.2 ============== UPGRADE URGENCY: MODERATE, introduced some bugfixes in the KEGG API and in imported PM parameters handling BUGFIXES: * KEGG API: quote the URLs to avoid HTTP error 400 on some systems * dphenome start: force negative parameters to zero for the clusterization * The parameters stored in the database are left unchanged * dape start: the maximum AV value in network reconstruction was still hard coded Version 0.14.0 ============== UPGRADE URGENCY: MODERATE, introduced some bugfixes in handling missing wells and in purging distinct replica BUGFIXES: * dphenome purge: replica id management bugfix * dphenome add/add-multi/add-dir have been changed accordingly * dphenome plot: handle missing strains * dphenome: handle missing wells in mutation experiments * dphenome: typo, "Sulfur" should be "Sulphur" Version 0.13.0 ============== UPGRADE URGENCY: HIGH, pangenome combined analysis have been changed to account for distant species The pathway maps and combined analysis have been remodulated to consider also the presence of genes that are not orthologs but share the same metabolic function (i.e. KEGG reaction ID). There are virtually no changes for pangenomic analyses on closely-related strains, as their metabolic variability is mainly based on orthology relationships. The KEGG reaction IDs that are common to all strains are labelled as 'conserved', as opposed to the 'variable' ones. FEATURES: * dgenome stats/export: new outputs and graphs for conserved/variable reactions * dape map: new pathway coloring policies for pangenomes * Blue boxes means conserved reaction * Orange boxes mean variable reactions (color intensity proportional to strains count) * dape start: consider conserved/variable reactions in the analysis * -n option removed BUGFIXES: * dphenome: the commands that need phenotypic data will exit if no PM plate is available Version 0.12.0 ============== UPGRADE URGENCY: MODERATE, crashes on CentOS have been fixed and new commands/options are available FEATURES: * dphenome trim: new command to cut all PM data to the same time step * The 'area' parameter is greater in longer PM curves, even if the curve shape is equal * dphenome purge: new purging policy ('replica') to remove a distinct replica by its ID * dape, dgenome, dphenome start: new option (-y) to fetch kegg data with unstable connections BUGFIXES: * dphenome start: support older scipy versions * dphenome start: support newer scikits.learn versions * dape, dgenome, dphenome start: correctly exit when kegg data fetch fails Version 0.10.0 ============== UPGRADE URGENCY: HIGH, in some cases there were crashes during curve parameters estimation FEATURES: * dphenome start: new option to plot an elbow test (-e) * dgenome export: more details in the reactions files BUGFIXES: * dphenome start: avoid crashes when estimating plateau and lag time for PM curves * dape map: reduced risks of crashes when downloading KEGG maps Version 0.9.6 ============= UPGRADE URGENCY: LOW, only a few improvements for BBH searches FEATURES: * dgenome start: two new options to modulate the BBH search * -m to change the Blast scoring matrix (i.e. BLOSUM62) * -e to change the Blast E-value threshold * Slightly faster threads Version 0.9.4 ============= UPGRADE URGENCY: LOW, only a style fix has been added BUGFIXES: * dphenome rings: thicker categories legend Version 0.9.2 ============= UPGRADE URGENCY: LOW, the bug fixed in this version affects only users importing curve parameters from other softwares BUGFIXES: * dphenome start: bugfix when considering parameters with NAN values in all wells Version 0.9.0 ============= UPGRADE URGENCY: HIGH, some compounds were discarded in network reconstructions FEATURES: * dape start: new outputs to higlight relevant genome/phenome combinations * pangenome: high genomic/phenomic variability are highlighted * mutants: high phenomic variability in pathways w/ mutated genes are highlighted * single organisms: all the genomic/phenomic data is showed BUGFIXES: * dape start: some compounds (w/ AV=0) were excluded from the network construction * dape start: discard the analysis on the general pathways * database: some queries have been optimized for pangenome analysis * Other small bugfixes Version 0.8.2 ============= UPGRADE URGENCY: HIGH, the curve parameters extraction and clustering have been improved FEATURES: * dphenome start: slight changes to parameters calculation * the smoothing algorithm is now less aggressive * the recursive attempts on curve parameters estimation are reduced two * dphenome start: clustering improvements * the curve parameters are normalized prior to clusterization, in order to give all the parameters the same weight * added a new option (-n) to set the number of clusters (and therefore tha maximum AV) BUGFIXES: * dape start: give the right number of distinct reaction IDs in the metabolic network tables * Other small bugfixes and speed improvements Version 0.8.0 ============= UPGRADE URGENCY: LOW, if you work exclusively with DuctApe and use the standard bacterial PM plates FEATURES: * dphenome import-plates: the user can now analyze even custom plates * the table in test/input/newplate.tsv contains an example * dphenome add/add-multi: PM data analyzed by different programs (i.e. opm) can now be imported * YAML/JSON file formats are supported * Curve parameters calculated by other programs are considered: they may be overridden by using dphenome start -r * dgenome export: PM data can now be exported in YAML/JSON format BUGFIXES: * dape init: if something goes wrong during DB creation raise an error * Other small bugfixes Version 0.7.0 ============= UPGRADE URGENCY: MODERATE; many features have changed, especially for the genomic analysis FEATURES: * Reactions copy number is no more analyzed: for pangenome and mutants experiments the program now focus on exclusive and distinct reaction IDs * dgenome stats/export, dape start/map have been changed accordingly * dgenome annotate: new command to transfer kegg annotations inside an orthologous group * dgenome deannotate: revert the annotation transfer * Added a test directory with a simple script to test the three modules BUGFIXES: * PM data: prevent crashes in the clusterization step * PM data: prevent crashes when parsing csv files * kegg: final fix to handle correctly HTMLs from kegg * DB: speed improvements on big pangenomes * Other small bugfixes Version 0.6.2 ============= UPGRADE URGENCY: HIGH; in some cases the KEGG html maps could not be retrieved correctly kegg: BUGFIX: the preparation of some kegg html maps resulted in failures dgenome start: FEATURE: roughly 3X speed improvements on BBH pangenome calculation, thanks to in-memory Blast searches Version 0.6.1 ============= UPGRADE URGENCY: HIGH; in some cases the KEGG data could not be retrieved correctly kegg: FEATURE: an improved connectivity check BUGFIX: the insertion in the database of some kegg maps resulted in failures dgenome start: BUGFIX: in a corner case the pangenome calculation was performed twice Version 0.6.0 ============= UPGRADE URGENCY: HIGH; the control signal subtraction has been fixed on PM plate PM04A dgenome start FEATURE: roughly 3X speed improvements on BBH pangenome calculation dphenome zero BUGFIX: the control signal subtraction has been fixed on PM plate PM04A dphenome export BUGFIX: the table with the average AV shouldn't contain the column "replica" kegg: FEATURE: fetch also the EC numbers for the Kegg reactions BUGFIX: add some spaces when a tag retrieved with the "info" request spans multiple lines dape start BUGFIX: fixing again the compound duplication bug when creating metabolic networks Version 0.5.4 ============= UPGRADE URGENCY: HIGH; dphenome users may be unable to load some PM data dphenome FEATURE: import/export plates data BUGFIX: biolog parser shouldn't fail on unknown plates BUGFIX: handle new possible plate names some plates names dape FEATURE: add-multi lets the user to provide multiple orgID Fixed some typos dape/dgenome/dphenome start: FEATURE: added options to skip kegg map and phenomics parameters calculation kegg: FEATURE: internet connectivity is checked before the kegg mapping is performed Version 0.5.3 ============= dape start FEATURE: added AV color to KEGG compounds (GML graph output) BUGFIX: avoid duplicated nodes in the metabolic graphs Version 0.5.2 ============= dape start BUGFIX: pangenome network analysis improvements dphenome add BUGFIX: handle csv files with a different OrgID than expected BUGFIX: calculate the correct number of added plates dphenome plot FEATURE: added the ability to generate plots in svg formats (-s option) Version 0.5.1 ============= BUGFIX: The net file was absent from the kegg module Version 0.5.0 ============= First release