As one part of the pipeline, I'm using MACS2 for MeRIP-seq differential peak calling.
First I used "callpeak" to generate a total of four bdg files (cond1 input, cond1 IP, cond2 input, and cond2 IP).
Then I used "bdgdiff" with these bdg files to find differentially enriched peaks.
I would like to use some statistics of differentially enriched peaks (e.g., fold changes, p/q-values). Since "bdgdiff" output BED files only provide likelihood ratio, I am currently comparing "callpeak" output and "bdgdiff" output to annotate overlapping peaks.
My criterion is to annotate "callpeak" peaks with cond1/cond2/common enrichment, when there's any "bdgdiff" peaks overlapping more than 50% of the "callpeak" peak region. I found a very small number of overlapping peaks: for example, callpeak=25000 peaks VS bdgdiff=200 peaks from three BED files. I've tweaked "bdgdiff" parameters (cutoff, min-len, and max-gap) but it does not seem to help.
My questions are following:
1) Is there any other way to calculate fold changes directly from "bdgdiff" output?
2) Is there any other MACS2 tool to help this problem? "bdgpeakcall" seems to give a similar result to "callpeak".
Hi,
As one part of the pipeline, I'm using MACS2 for MeRIP-seq differential peak calling. First I used "callpeak" to generate a total of four bdg files (cond1 input, cond1 IP, cond2 input, and cond2 IP). Then I used "bdgdiff" with these bdg files to find differentially enriched peaks. I would like to use some statistics of differentially enriched peaks (e.g., fold changes, p/q-values). Since "bdgdiff" output BED files only provide likelihood ratio, I am currently comparing "callpeak" output and "bdgdiff" output to annotate overlapping peaks.
My criterion is to annotate "callpeak" peaks with cond1/cond2/common enrichment, when there's any "bdgdiff" peaks overlapping more than 50% of the "callpeak" peak region. I found a very small number of overlapping peaks: for example, callpeak=25000 peaks VS bdgdiff=200 peaks from three BED files. I've tweaked "bdgdiff" parameters (cutoff, min-len, and max-gap) but it does not seem to help.
My questions are following: 1) Is there any other way to calculate fold changes directly from "bdgdiff" output? 2) Is there any other MACS2 tool to help this problem? "bdgpeakcall" seems to give a similar result to "callpeak".
Thanks in advance!
Taehyung