Hello,

I have successfully ran DiffBind on K4me3 and K9me3, so thanks for the very clear Vignette! However, I probably missed a key information : I am wondering if the peaks I'm feeding DiffBind should be FDR filtered or not. I called peaks using MACS3 which has a default FDR filtering <0.05 and obtained between 3000 (K9me3) and 6000 (K4me3) broad/narrow peaks. So the question is : should I call peaks with MACS3 without the default FDR filtering, before inputing the peaks to DiffBind?

Thank you,

Rita

Generally, you should use FDR filtered peaks.