Venn Diagrams May Indicate Erroneous Statistical Reasoning in Transcriptomics.

ABSTRACT: A common application of differential expression analysis is finding genes that are differentially expressed upon treatment in only one out of several groups of samples. One of the approaches is to test for significant difference in expression between treatment and control separately in the two groups, and then select genes that show statistical significance in one group only. This approach is then often combined with a gene set enrichment analysis to find pathways and gene sets regulated by treatment in only this group. Here we show that this procedure is statistically incorrect and that the interaction between treatment and group should be tested instead. Moreover, we show that gene set enrichment analysis applied to such incorrectly defined genes group-specific genes may result in mislea

SUBMITTER: Weiner J 

PROVIDER: S-EPMC9046926 | biostudies-literature | 2022

REPOSITORIES: biostudies-literature