how to merge peaks when finished IDR consistency test in ATAC-seq data?

Question

[hsu]

hi, I have 3 biological replicates of ATAC-seq data, named a,b,c. I want to find some consistent peaks among these replicates. And I apply idr ---compare ab,ac,bc respectively. Thus, I get three bed files, but I don't know how to merge these files to get a consistent peak file of 3 replicates. Could you help me solve this problem? Thanks a lot!

[ATpoint]

How about bedtools merge? Or use a peak caller that takes replicates right away such as Genrich. DiffBind (Bioconductor package) has some functions to merge peaks as well.

[hsu]

I got bedfiles from idr software. I want to merge these consistent peaks but not the original peaks. bedtools merge may not help....

[ATpoint]

Please use `ADD COMMENT/REPLY´ for comments, not the answer field, otherwise the thread quickly becomes unorganized.

I do not quite follow. So you should have three peak lists, one from each sample, and you ran IDR 3 times, right, so comparing ab, bc, ac, is that correct? That should give you three "IDR" lists. Now you could merge them with bedtools merge like:

cat ab_idr.bed ac_idr.bed bc_idr.bed | sort -k1,1 -k2,2n | bedtools merge -i - > merged.bed

Does that make sense here?

[zechen]

hi,I had the same problem,did you already have a solution?Looking forward to your reply!Thanks a lot!