]> Statistics Ontology This module provides terms and relations for representing statistical models and analyses that quantify variation, effects, and quality across experimental, process development, testing, and historical data. It integrates core concepts from STATO within the IOF framework to support design of experiments, regression and variance analysis, response surface methodology, and statistical evaluation used for process characterization, monitoring, and validation. http://opensource.org/licenses/MIT

covariate variable https://spec.industrialontologies.org/ontology/biopharma/Statistics/ dissolved oxygen concentration recorded throughout each cell culture run and specified as a covariate when feed strategy is a factor and harvest titer is the dependent variable; actual culture temperature recorded throughout each bioreactor run and specified as a covariate when temperature is not deliberately varied; initial viable cell density specified as a covariate when culture temperature and feed strategy are factors; seed culture age at inoculation recorded for each bioreactor run and specified as a covariate in an analysis of culture growth and productivity

http://purl.obolibrary.org/obo/OBCS_0000040

CovariateVariable(x) → Variable(x)

true

variable representing an observed rather than deliberately assigned characteristic that is specified for accounting for variation in a dependent variable separately from variation associated with the factors under study

There are insufficient constructs to create a set of necessary and sufficient conditions.

if x is a 'covariate variable' then x is a 'variable'

data preparation process https://spec.industrialontologies.org/ontology/biopharma/Statistics/ removing duplicate records from analytical data; converting measurement results to common units; harmonizing timestamps from equipment and laboratory systems; joining experimental-run metadata with analytical results; imputing missing observations according to a specified method; encoding categorical variables for statistical modeling; applying logarithmic transformations to skewed variables; standardizing variables before multivariate analysis

Data preparation involves applying one or more data transformations to convert raw or heterogeneous data into a consistent, usable form. It typically includes steps such as handling missing values, correcting errors, standardizing formats, or restructuring data for analysis, integration, or reporting.

DataPreparationProcess(x) → ComputingProcess(x)

true

computing process in which input data are cleaned, organized, or otherwise transformed to make them suitable for subsequent analysis or use

There are insufficient constructs to create a set of necessary and sufficient conditions.

if x is a 'data preparation process' then x is a 'computing process'

explained variance https://spec.industrialontologies.org/ontology/biopharma/Statistics/ coefficient of determination that represents the variance explained by a linear model; eta-squared that represents the variance explained in an ANOVA analysis

https://www.statology.org/explained-variance/

ExplainedVariance(x) → Statistic(x)

true

statistic representing the portion of the total variance in a dependent variable that is accounted for by one or more variables included in a statistical model

There are insufficient constructs to create a set of necessary and sufficient conditions.

if x is an 'explained variance' then x is a 'statistic'

exploratory data analysis https://spec.industrialontologies.org/ontology/biopharma/Statistics/ computing summary statistics such as mean, variance, or correlation coefficients; plotting histograms, boxplots, or scatter matrices to visualize data distributions and relationships; identifying missing values or outliers in experimental results before modeling

https://www.iso.org/obp/ui/en/#iso:std:iso-iec:22989:ed-1:v1:en:term:3.2.6 and https://www.ibm.com/think/topics/exploratory-data-analysis

Exploratory Data Analysis involves determining and describing features through summary statistics, visualization, and other descriptive techniques. It supports understanding of data structure, relationships, outliers, and quality before applying formal modeling or hypothesis testing.

ExploratoryDataAnalysis(x) → PlannedProcess(x)

true

planned process used to determine and analyze characteristics of input data in order to identify structure, trends, or irregularities

There are insufficient constructs to create a set of necessary and sufficient conditions.

if x is an 'exploratory data analysis' then x is a 'planned process'

outlier detection https://spec.industrialontologies.org/ontology/biopharma/Statistics/ applying a generalized extreme Studentized deviate test to replicate assay results to identify one or more potential outliers; applying an interquartile-range criterion to supplier-lot impurity results to flag atypical observations; applying robust multivariate-distance criteria to metabolite profiles to identify atypical culture samples; applying an isolation-based anomaly-detection algorithm to historical process data to flag observations for investigation

https://www.sciencedirect.com/topics/mathematics/outlier-detection#:~:text=Outlier%20detection%20is%20defined%20as,fraud%20detection%20and%20data%20cleaning.&text=How%20useful%20is%20this%20definition?

1) This class is intended to cover a broader range of outlier detection approaches than those currently represented in STATO, including both hypothesis-based statistical tests and non-hypothesis or algorithmic methods for identifying outliers. 2) Outlier detection is distinct from evaluation against a material specification, statistical control-chart rules, or an expected process trend. An observation may be flagged as a potential outlier without being out of specification, out of control, or out of trend, and those classifications do not by themselves establish that the observation is an outlier.

DataTransformation(x) ∧ ∃y (achievesPlannedObjective(x, y) ∧ OutlierDetectionTestingObjective(y)) → OutlierDetection(x)

true

computing process in which input data are evaluated for the presence of outliers according to a detection criterion

There are insufficient constructs to create a set of necessary and sufficient conditions.

if x is a 'data transformation' and x 'achieves planned objective' some 'outlier detection testing objective' then x is an 'outlier detection'

response surface plot https://spec.industrialontologies.org/ontology/biopharma/Statistics/ three-dimensional surface showing predicted product titer across combinations of culture temperature and pH while feed rate is held fixed; contour plot showing predicted galactosylated-species abundance across coded manganese and zinc concentrations; response surface showing predicted chromatography yield as a function of load density and residence time at a fixed buffer pH

https://www.jmp.com/en/statistics-knowledge-portal/design-of-experiments/response-surface-methodology and https://wolfson.huji.ac.il/purification/PDF/Others/GE_DOE_in_Protein_Production_and_Purification_Handbook.pdf

A response surface plot is typically constructed from the predictions of a fitted regression model, most often a second-order (quadratic) polynomial, derived from data obtained in a designed experiment such as a central composite or Box–Behnken design. The figure illustrates how the predicted value of the dependent variable changes as the independent variables vary and may be presented as either a three-dimensional surface or a two-dimensional contour projection. Such plots facilitate the interpretation of main, interaction, and curvature effects and assist in locating regions of optimal or critical response.

ResponseSurfacePlot(x) → Figure(x)

true

figure that depicts the predicted values of a dependent variable as a continuous surface over two or more independent variables

There are insufficient constructs to create a set of necessary and sufficient conditions.

if x is a 'response surface plot' then x is a 'figure'

contains values for variable https://spec.industrialontologies.org/ontology/biopharma/Statistics/ time-series dataset used in parameter fitting that holds pH measurements for the covariate variable denoting pH; design matrix providing the coded temperature levels (–1, 1) for twelve experimental runs of the temperature factor variable; regression dataset providing z-score-normalized substrate concentration values for an independent variable; single-run dataset recording the product titer value (measured or predicted) for the dependent variable representing product yield

relation from an information content entity to a variable that holds when the information content entity contains one or more values associated with that variable

quantifies statistical association of https://spec.industrialontologies.org/ontology/biopharma/Statistics/ a correlation coefficient that quantifies statistical association of a temperature variable and a growth-rate variable; a mutual-information value that quantifies statistical association of a transcriptomic profile and a proteomic profile

relation from a measure of association to an information content entity involved in the statistical relationship quantified by that measure