q-value > 1 in STAMP

Question

[zhangdengwei]

Hi all,

I am using STAMP, a powerful tool in statistics, to determine which bacteria is differential in my sample. "BH" was adopted to correct the p-value, but the q-value is > 1. It is, in theory, impossible. Does anyone know what happens? Thanks very much.

genus p-values p-values (corrected)

g__Collinsella 0.931588858 1.054628896

g__Oscillibacter 0.914789867 1.05552677

[Kevin Blighe]

In case there is no response, you may consider contacting the developer. E-mail available here: https://beikolab.cs.dal.ca/software/STAMP_FAQs

Answer 1

[zhangdengwei]

thanks. I have got a reply from the developer. STAMP uses the following definition to calculate BH q-value: q-value = p-value*(total number of hypotheses tested)/(rank of the p-value) These are not restricted to be <1.0.

[Kevin Blighe]

Thank you. I have moved this to an answer.