Different peak sites from .xls and .narrowPeak, macs2 output

Question

[sh.kazempour94]

Hi all, I used macs2 callpeak in order to find the peaks in the ATAC-Seq data. The reads are paired-end and this is my command to get the peaks:

macs2 callpeak -t $TREATPATH -f BAMPE -n ${BASETREAT} -g mm --outdir $2/peaks --keep-dup all -B -q 0.01 --SPMR --shift 100 --extsize 200 --nomodel

I have a .bdg file which I converted to bigwig file for visualization. There are also .narrowPeak and .xls files. when I compared these two files for a specific region (with strong peaks according to bw file visualization, e.g. chr6:52,000,000-52,300,000), I realized that .xls file is missing some of the regions compare to .narrowPeak. I always supposed that these two files contain the same genomic regions with extra information in each of them, but apparently it's not the case. My question is that why these two files differ? What does macs2 do differently in generating the .xls file compared to .narrowPeak?

Than you,

Shiva.

Answer 1

[ATpoint]

Do yourself a favor and simply ignore this Excel file, the narrowPeak file is what matters. Maybe an Excel-like conversion error or something like this.

[sh.kazempour94]

Thank you for your reply. Then if I use .narrowPeak files for further analysis using DiffBind, I realized that some of the regions that are available in all sample are excluded from binding matrix. I asked the question with more details here. If you have any idea why this is happening, I would really appreciate your help. Thanks.