David Output Interpretation

Question

[TitoPullo]

I'm trying to understand bettern the output of an enrichment analysis done with DAVID . To each overrepresented term is associated a p-value and a Benjamini corrected p-value. Which one should be taken in account? It looks like that the Benjamini is the most correct one because it consider the multiple hypothesis problem, but then why show us also the normal p-value?

Answer 1

[Devon Ryan]

The main reason (that I can think of at least) to show the raw p-value is that DAVID may be testing a bunch of GO categories that you don't actually care about. If you only cared about a couple categories and were using DAVID as a convenient web-interface to get p-values for them, regardless of the other results, then having only the adjusted p-values would be undesirable. If you used DAVID for the normal "let's just see if any categories are enriched, hopefully they make biological sense" way, then pay attention to the adjusted p-value.