This is a test version of Biostars. For the public version, visit https://www.biostars.org.
umitools dedup and quality filtering

Hi,

I am using a pipeline on paired WGS data where the main steps are: 1) alignment 2) merge BAM across lanes 3) deduplication using umitools dedup 4) quality filtering MAPQ values 5) remove singletons

However, I am wondering if there are any best practices for swapping steps 3) and 4), ie. Quality filtering immediately after merging BAMs, then performing deduplication with UMIs. I have ran a few samples through these two alternate versions of the pipeline (setting the same seed for UMItools), and at a total read count level there is only a minimal difference (<1%).

Thanks

umitools filtering wgs quality deduplication
• 750 views • link
updated by i.sudbery i.sudbery • written by song • song

1 answer

The only dependence between umi-tools deduplication and MAPQ is that where multiple reads have the same UMI and same mapping location, the read with the highest MAPQ will be chosen. It might also, concievably affect the formation of UMI networks - for example if you had a location with UMIs AAAA, AAAT and AATT, when if AAAT was low MAPQ, filtering it might break the association of AAAT to AAAA, but I think this is likely uncommon unless the sample is highly over sequenced.

• 0 views • link
written by i.sudbery i.sudbery

Log in to answer this question.

More posts like this