[help] Can I use MethylDackel for diploid reference genome please?

[lizhe-gis]

Dear Devon,

Thank you so much for developing this great tool! My colleague recommended bwa-meth and MethylDackel to me.

Recently we developed a haplotype-resolved diploid human genome, where one copy is paternal and the other copy is maternal. I imagine if I map WGBS data to this diploid reference genome with bwa-meth, most reads will have secondary alignment with MAQ0 due to the high similarity of pat-/mat-genome. From reading the previous posts and the alignment flags, I got the impression that secondary alignment will be discarded (#25 (comment)), I am just wondering whether it is possible to set MAPQ >= 0 with -q and retain all reads with secondary alignment and only use the primary alignment (which I assume is randomly assigned?) for calculating the methylation percentage please?

Thank you very much!

Best Regards,
Zhe

[dpryan79]

Sorry for the late reply. You'll want the following options: -q 0 -F 3584. I think that should suffice for your needs.