Multiple sources of bias confound functional enrichment analysis of global -omics data

Abstract

Serious and underappreciated sources of bias mean that extreme caution should be applied when using or interpreting functional enrichment analysis to validate findings from global RNA- or protein-expression analyses.