Dear Bioc Users, I'm using the limma library for differential gene expression analysis. I have a question about how to choose the p-value. These are my contrasts: T1 - Control T2 - Control T3 - Control T4 - Control T5 - Control Then I applied: fit2 <- contrasts.fit(fit, contrast.matrix); eb <- eBayes(fit2) Now I should choose all those genes differentially expressed either by using topTable or decideTests. However I wouldn't like to choose the p-value arbitrarily. I did: results <- decideTests(eb, p.value=0.05, method="global", adjust.method="BH") I know it means I will get an FDR < 0.05 for the whole set. However I have no idea about which "method" (from the arguments) would be the right one... I'm still getting many genes differentially expressed, that's why I'm wondering whether this is correct or not. Thank for your help in advance. Yours, Avhena [[alternative HTML version deleted]]