outputing statistics about anchor peaks

[armenabnousi]

I'm running hichipper on some samples and I'm interested in the background corrected anchors that hichipper is finding (not the loops). I can see "anchors.bed" file and I also keep the temporary files.
The problem is I'm getting too many anchors (~400K anchors, spread over ~500K bins of size 5K), I'm using hg38 so that's almost 83% of the possible 5K bins.
Another problem is that anchors.bed file does not contain any statistical significance. When I'm looking at temporary files, the values in "allSamples_temporary_hichipperPeaks.bed" are all 0. But the "hichipper_qvalue.bdg.tmp" seems have different qvalues (between 0 to 103000, with median 1.4). Do you think this means I should change the parameters for "macs2 bdgpeakcall" in hichipper? Why am I getting all 0's in my hichipperPeaks.bed file?

Thanks!

[caleblareau]

Hi @armenabnousi, this is a good question, and I think that you're on the right track. The hichipper_qvalue.bdg.tmp indeed serves as input to macs2 bdgpeakcall (see line 116: https://github.com/aryeelab/hichipper/blob/master/hichipper/hicproHelper.py). I would imagine that taking this file and running the command with more stringent q value thresholds could be a sensible path forward and should minimally give you a score per peak to start interpreting. I'm not entirely sure why you are getting a bunch of zeros however...