Compute Cohen's f for All Gene Signatures Across a Categorical Variable
Source:R/Heatmap_Cohen.R
CohenF_allConditions.RdComputes Cohen's f effect sizes and corresponding p-values for all gene signatures using scores calculated
by multiple methods. The function first computes gene signature scores using CalculateScores with the "all"
option, flattens the results, and fits linear models using the specified variable to estimate effect sizes.
Arguments
- data
A data frame of gene expression data, with genes as rows and samples as columns.
- metadata
A data frame containing sample metadata. The first column should contain sample identifiers matching the column names of
data.- gene_sets
A named list of gene sets. For unidirectional gene sets, each element is a vector of gene names; for bidirectional gene sets, each element is a data frame where the first column contains gene names and the second column indicates the expected direction (1 for upregulated, -1 for downregulated).
- variable
A string specifying the categorical variable in
metadataused to model the gene signature scores.
Value
A named list where each element corresponds to a gene signature. Each signature element is a list with three components:
- CohenF
A data frame where rows are scoring methods and columns are the variable used in the linear model (usually one column), containing the computed Cohen's f effect size.
- PValue
A data frame of the corresponding raw p-values from the linear model for each method.
- padj
A data frame of adjusted p-values (Benjamini-Hochberg method) across signatures and contrasts, per method.