No effect on output called peaks by tuning nomodel+shift+extsize

[linpei26]

Dear Dr. Liu

I have two conditions for my ChIP-Seq experiment. We observed ~10,000 called narrow peaks in 1 condition but ~1,000 in the other (FDR < 0.1)

So we asked if tuning some parameters of MACS2 (version 2.2.7) may affect the output called peaks. I tried some different values of extsize with setting shift = 0 and used --nomodel. Then I also try different values of shift.

All results found same number of called narrow peaks (FDR < 0.1)

Is this a very expected outcome? Are there any para should be examined first when users found too many or too few peaks are called?

Thanks!

Pei