Hi all,

I'm going to calculate co-occurrence statistics among several TF peaks. Homer document says mergePeaks uses hypergeometric distribution to calculate p value for overlap but the program only takes genome size as background.

1. How they establish the hypergeometric model? Does mergePeaks ignore continuity of peaks and only use base as unit?
2. If I want to get co-occurrence statistics in TSS, intergenic, TTS and introns respectively, what background should I provide? Is there any existing tool for this task?

Thanks!