haowenz / chromap

Fast alignment and preprocessing of chromatin profiles
https://haowenz.github.io/chromap/
MIT License
187 stars 20 forks source link

Chromap miss some region when deal with Hetero Chromatin #141

Open ming1211 opened 11 months ago

ming1211 commented 11 months ago

Thanks for Chromap wihch is really really helpful especially when dealing with very big dataset. But I found that Chromap miss some region when deal with Hetero Chromatin As shown in the figure, the bw generated by Bowtie2 is Blue while the chromap generated are in orange. Have you encounter this situation before? Are there anything I can do to improve it? image

Thanks in advance!

haowenz commented 11 months ago

Did you run QC on Bowtie2 mappings? Chromap performs QC and removes PCR duplicates and mappings with low MAPQ.

mourisl commented 11 months ago

Chromap internally automatically conducts MAPQ filtering, which is a standard step in data postprocessing. For the multimapped reads, their MAPQs are always 0 and then filtered. For the heterochromatin region, it seems to be highly repetitive region, therefore reads will be multimapped. If you need those reads, you can add the "-q 0" option.

ming1211 commented 11 months ago

I see! Thanks for your reply! It's very helpful, I will try -q 0 and update with you! Thanks again!

Li Song @.***> 于2023年10月6日周五 18:01写道:

Chromap internally automatically conducts MAPQ filtering, which is a standard step in data postprocessing. For the multimapped reads, their MAPQs are always 0 and then filtered. For the heterochromatin region, it seems to be highly repetitive region, therefore reads will be multimapped. If you need those reads, you can add the "-q 0" option.

— Reply to this email directly, view it on GitHub https://github.com/haowenz/chromap/issues/141#issuecomment-1750991908, or unsubscribe https://github.com/notifications/unsubscribe-auth/A4AE3ZHVAJUIAUHV53NNUR3X6ATNNAVCNFSM6AAAAAA5VNLRZGVHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMYTONJQHE4TCOJQHA . You are receiving this because you authored the thread.Message ID: @.***>