To celebrate the release of GATK 4.0, we are giving away free credits for running the GATK4 Best Practices pipelines in FireCloud, our secure online analysis portal. It’s first come first serve, so sign up now to claim your free credits worth $250. Sponsored by Google Cloud. Learn more at

Reference genotype quality in presence of conflicting reads

Given the following line from a VCF file:

Supercontig_1.1 308 . T A . PASS AC=0;AF=0;AN=2;BaseQRankSum=-1.932;ClippingRankSum=0;DP=99;ExcessHet=3.01;MQ=43.26;MQRankSum=-2.382;ReadPosRankSum=1.77;VariantType=SNP GT:AD:DP:RGQ 0/0:52,4:56:1

I note that despite there being 52 reads passing filters for the reference genotype, the reference genotype quality is still only 1. Is RGQ affected by the presence of reads indicating a possible variant (4 in this case)? So the low RGQ score in this case reflects uncertainty over whether this position really is reference call (T/T), or if it might be a variant (A/A or A/T or T/A).

If I was being super strict about only including highly certain positions in my analysis would you recommend that I assign this position a missing genotype because I can't really be sure what it is?

Best Answer


  • SheilaSheila Broad InstituteMember, Broadie, Moderator


    Did you use the GVCF workflow to get the final VCF? If so, can you post the GVCF record for the site? Please also post IGV screenshots of the original BAM file and bamout file.


  • simono101simono101 London, UKMember
    edited September 2016

    Hi @Sheila,
    I did use the gVCF workflow. Here is the gVCF record for that site:

    Supercontig_1.1 308 . T A,<NON_REF> 0.01 . BaseQRankSum=-1.932;ClippingRankSum=0.000;DP=99;ExcessHet=3.0103;MLEAC=0,0;MLEAF=0.00,0.00;MQRankSum=-2.382;RAW_MQ=185235.00;ReadPosRankSum=1.767 GT:AD:DP:GQ:PL:SB 0/0:52,4,0:56:1:0,1,2314,169,2326,2494:52,0,4,0

    And attached are two screenshots centered on position Supercontig_1.1 308 for the bamout from HC and the original BAM.

    Many thanks,


Sign In or Register to comment.