fisher_enrichment: Fisher Exact Test of gene set enrichment with output to Excel

In jdreyf/ezlimma: Streamlines and extends limma package

Fisher Exact Test of gene set enrichment with output to Excel

Description

Test enrichment of one or more vectors of significant genes against the universe of genes in feat.tab per gene set using fisher.test. It returns a data frame with statistics per gene set, and can write this to Excel. The Excel file links to CSV files, which contain statistics per genes in a set.

Usage

fisher_enrichment(
  sig.set,
  G,
  feat.tab,
  name = NA,
  adjust.method = "BH",
  min.nfeats = 3,
  max.nfeats = 1000,
  pwy.nchar = 199
)

Arguments

sig.set	Named list of length one whose sole element is a vector of significant gene IDs matching rownames(feat.tab).
G	Gene set list as returned from read_gmt.
feat.tab	Table of feature (e.g. gene) statistics that the Excel table can link to.
name	Name for the folder and Excel file that get written. Set to NA to avoid writing output.
adjust.method	Method used to adjust the p-values for multiple testing. Only for mroast.
min.nfeats	Minimum number of features (e.g. genes) needed in a gene set for testing.
max.nfeats	Maximum number of features (e.g. genes) needed in a gene set for testing.
pwy.nchar	Numeric maximum number of characters allowed in pathway name.

Details

Pathway (i.e. gene set) names are altered to be valid filenames in Windows and Linux. Numeric columns are rounded to 8 significant figures.

Value

Table of pathway statistics with the number of genes from feat.tab in the pathway, the number of these genes that are in sig.set, the p-value, and the adjusted p-value from the one-sided Fisher exact test.

Examples

 G = list(s1=list(name = "s1", description=NULL, genes=letters[3:10]), 
   s2=list(name = "s2", description=NULL, genes=letters[2:99]))
 feat.tab <- matrix(rnorm(18), ncol=3, dimnames=list(letters[1:6], paste0("col", 1:3)))
 fisher_enrichment(sig.set = list(A=letters[1:3]), G = G, feat.tab = feat.tab)

jdreyf/ezlimma documentation built on Sept. 16, 2024, 2:08 a.m.