DAVID: p-value and Benjamini correction interpretation

Question

[Mel]

Hi everybody!

I submitted a list of gene to DAVID Functional annotation chart tool. I decided to used EASE (p-value) cutoff of 0,05 (from the DAVID publication on Nature Protocols in 2009) and selected Benjamini correction. Now I have problems with results interpretation. Is the Benjamini value the (i/m)Q value? So if it is bigger than my p-value, is my term significantly enriched? Or should I interpret the Benjamini value like a new p-value corrected already and is it significant if it is <0.05? (It looks strange for a multiple comparison tests).

Thank u

[kehoe]

Hi Mel,

I was wondering about your chosen cut-off. Why did you choose to change from the default EASE 1.0 to 0.05 instead? I am trying to understand how the cut-off works to apply to my data. There doesn't seem to be much information about the exploration of this so how did that threshold work out for your data?

Answer 1

[arta]

Benjamini correction is your false discovery rate and it is your adjusted p-value. So you should forget about your p-value after correction. So your test is significant if your adjusted p-value is smaller than criteria (such as 0.05 or 0.01). If you want to more about multiple testing, you can check here

Answer 2

[Devon Ryan]

Corrected p-values should be interpreted exactly the same as regular p-values (they're just corrected for multiple testing). Whether an adjusted p-value < 0.05 is significant is up to you (as is the case for unadjusted p-values).