getChrCGFromCools Assumption

[helengracehuang]

The getChrCGFromCools function assumes the same interactions being observed in all samples, even if some have a value of 0. Is there any good way to standardize all the cooler files (essentially filling in the zeros)? I converted .hic to .cool and each file have slightly different list of interactions (probably because bins with 0s are omitted), so I can't just use one bin.ixns for all replicates.

[Gemma-Zhang-326]

Hi, Dr Huang, The same issue occurs to me, I wonder how you solve this problem, or which tool you choose to use later on?

[cafletezbrant]

As of now, this is not something we have developed specifically. I think, from browsing https://cooler.readthedocs.io/en/latest/cli.html#cooler-cload, that if you for example have a table with columns indicating matrix cells and also the read count in each cell (this is the cload pairs section), and augment this table to include cells with 0 read counts, then cooler will not delete the 0-valued cells. In this way you could standardize cooler files.