how to keep multi-mapped paires for HiC data.

[ming1211]

Dear you,

Thanks for the continued updates, the high speed of chromap is really helpful for giant genomes like wheat!
I'm dealing with Wheat Hic data, because of the similarity of the genome, I'd like to keep multi-mapped pairs.
Compared to Hic-Pro which useses bowtie2 can set RM_MULTI = 0 to achieve it, Chromap I don't know which parameter can be set for this. (if set the mapQ as 0, will it make the results containing many results with unexpected mismatches?)

and also, will you take the enzyme at the clip sites into consideration for next version of chromap?

Thanks for your effort!

[mourisl]

Most of the MAPQ 0 is due to multi-mapped pairs, so I don't think it will create too many unexpected mismatches. You can reduce the value of -e if this is a concern.

I think Hic-Pro may use some fields in the BAM file to know how many places a read is mapped to, or is it based on sort by read name? If it is based on read ids, you may need to add the option like "-n 2" to make Chromap output two mappings for multi-mapped reads.