What is the bin size for Bamcompare?

Question

[Emily]

If I'm working on C elegans chipseq data, for bamcompare is it ok to choose a bin size of 50? Is the choice of bin size determined by size of genome? Particularly smaller genome choose a smaller bin size to get better resolution? Or is it determine by size of peak like if it is broad peak use larger bin size?

Answer 1

[ATpoint]

It is the number of adjacent bases that are aggregated (binned) into a single value. I find binning not optimal, and would always set this to 1. Bigwigs on the genome browser look ugly with binning, and you could later always bin basepair-resolution bigwigs (binsize of 1) in downstream analysis if necessary. Binning saves space, that's it. I don't like that option.