I'm curious as to how MACS2 assigns a p-value due to the following problem I'm having right now.
I've ran MACS2 on some ATAC-seq data with the default parameters and the output has an assigned pValue associated with each peak. However, I wanted to hone in on the peaks we have most confidence in so I reran MACS2 using a pvalue cutoff of 1e-5 (parameter: -p 1e-5).
I was under the assumption that changing the pval cutoff would just subset my original macs2 run (default parameters) with pvalues lower than 1e-5. However that is not the case and even many regions that overlap between the MACS2 runs have different associated pvalues with said region.
Can anyone explain to me how pValue is calculated for MACS2 and how changing the cutoff would affect my results this way?
Thanks!
