Hi everyone,
I am working with DiffBind to analyze ATAC-seq data, specifically using peaks generated by HMMRATAC. I ran the dba.count() function with the following parameters:
dba_obj <- dba.count(dba_obj, bUseSummarizeOverlaps = TRUE, summits = 75)
However, I am encountering the following error:
Error: No sites have activity greater than filter value.
I tried increasing the summits parameter, but the error persists. Interestingly, when I set bUseSummarizeOverlaps = FALSE, the error does not appear, and DiffBind proceeds with counting without any issues.
My questions are:
Does setting
bUseSummarizeOverlaps = FALSEsignificantly impact the accuracy or validity of my results?I understand thatSummarizeOverlapsgenerally provides more stringent counting, but I’m unsure how crucial this setting is for ATAC-seq data analysis, particularly with peaks from HMMRATAC.Are there any alternative ways to troubleshoot or modify settings that might resolve this error without disabling
bUseSummarizeOverlaps?
Any guidance on whether or not to keep bUseSummarizeOverlaps enabled, or on alternative approaches to overcome this issue, would be greatly appreciated.
Thank you for your help!