Mutect2 output with samples with different read groups

quaglieri_a16quaglieri_a16 MelbourneMember

Hi all,
I am running Mutect2 with a tumour/normal match in RNA-Seq. Here is my command:

gatk -T MuTect2 -R ./genome.fa --read_filter Mappin
gQuality --min_mapping_quality_score 20 --max_alternate_alleles 4 -kmerSize 15 -kmerSize 30 -I:tumor ./9_Recal.reorderedDupl.rg.split.bam -I:normal ./8_Recal.reorderedDupl.rg.split.bam --dbsnp ./dbsnp_144.hg38.vcf.gz -L ./genes_for_variant_calling.bed -o ./9_somatic_snvs_indels.vcf -log ./9_somatic_snvs_indels_log

In the output (I am showing here output opened in R with the VariantAnnotation package) I get the following data for the genotypes:

                                        8_L003       8_L004     8_L007     8_L008        9_L003     9_L004       9_L007    9_L008

chr1:36466762_A/G "0/0" "." "." "." "0/1" "." "." "."
chr1:36468081_A/G "0/0" "." "." "." "0/1" "." "." "."
chr1:36475819_A/G "0/0" "." "." "." "0/1" "." "." "."
chr1:36476164_A/G "0/0" "." "." "." "0/1" "." "." "."
chr1:36476671_C/CT "0/0" "." "." "." "0/1" "." "." "."
chr1:36476745_T/C "0/0" "." "." "." "0/1" "." "." "."

I get the same strange output for many other samples. It is like the variants are found in a subset of reads related to one one read group. All the variants are like this. Shall I get really worried about my data or Mutect2 does some comparisons that I am not aware of?

Thanks for your help,


Best Answer


  • SheilaSheila Broad InstituteMember, Broadie, Moderator admin
    Accepted Answer

    Hi Anna,

    Can you tell us which version of GATK you are using? Can you also post the records straight from the VCF here? You can just post the first few records in the VCF.


  • quaglieri_a16quaglieri_a16 MelbourneMember

    Hi Sheila,

    Thanks for your reply.

    I am really sorry for my long delay but I hadn't receive a notification about this and I forgot to reply.
    I tried to reproduce the problem using the same file and now it looks fine, it does not stratify based on Read Groups.

    vcf <- readVCF("9_somatic_snvs_indels.vcf")

                                TUMOR NORMAL

    chr1:36466130_A/G "0/1" "0/0"
    chr1:36466762_A/G "0/1" "0/0"
    chr1:36467559_T/C "0/1" "0/0"
    chr1:36467591_A/G "0/1" "0/0"

    I am suspecting that something has changed into the VariantAnnotation package.

    Thanks for your help,


  • SheilaSheila Broad InstituteMember, Broadie, Moderator admin

    Hi Anna,

    Thanks for getting back over a year later! :smile: You can set the notifications using this article. I think your issue has been resolved, so this is great.


Sign In or Register to comment.