does Mutect2 fit for deepseq?

sergey_ko13sergey_ko13 Member
edited September 2018 in Ask the GATK team

Dear GATK team and forum members,

After some good experience with raw reads pre-processing guidelines and mutect2 we wonder if we could use it for our deepseq data.
There must be tons of differences in nature of reads.

It is clear that Picard's MarkDuplicates step is omitted as all reads in a locus have same start and that --max-reads-per-alignment-start of Mutect2 should be disabled to keep all reads in focus of test.

But what are other possible limitations of WES/WGS-designed workflows for our case?
How BQSR behaves with so many variants happening at same sequencing cycle (variant reads have same start)? Does mutect2 fit for such purposes at all?

We have gene pannel with 1350 amplicons (~170 bases each) sequenced on MiniSeq (2x150b) with 1k depth.

We appreciate any advise or suggestion



  • SheilaSheila Broad InstituteMember, Broadie, Moderator admin

    Hi Sergey,

    People have reported issues with reassembly based variant callers on amplicon data. Have a look here. However, the team is interested in fixing any reassembly related issues, so please let us know if you do find anything unusual.


Sign In or Register to comment.