What is the difference between bed files from bigWig or bam, and bed files of narrow/broad peaks?

Question

[1888]

Hi,

Can someone point to procedures on how to get a bed file from a bigWig file, that somewhat reflects broad/narrow peak regions called in the same cell e.g. from Encode (if this is possible)?

I know it is possible to get a bed file from a bigWig file in Encode for example using the utility 'bigWigToBedGraph', but the resulting bed regions are very different from the regions in the bed files in broad/narrow peaks - is there a way to converge to similar regions?

Thank you!

Answer 1

[Devon Ryan]

bigWig files are generally whole-genome coverage tracks, so you would need to call peaks on them. You can convert them to bedGraph format and then input that into MACS2 bdgpeakcall. Note however that you may have to manually play around with the parameters quite a bit to get reasonable peak calls out of that. In general it's preferable to try and find the original BAM files from which the bigWig files were generated and just give those to a peak caller, since that's what the peak callers are written to work with.