0 AD/DP for called variant in HaplotypeCaller

Follow

LucaB

• April 05, 2024 10:23

Hi, I am using GATK v.4.5.0.0. and calling variants with HaplotypeCaller.

While investigating the minimum variant allele frequency called by HaplotypeCaller, I noticed that some variants are called as heterozygous or homozygous while reporting in the FORMAT/AD 0 reads supporting the alternative allel and a drastically differnet FORMAT/DP value with respect to what observed in the BAM files. Here I display an example (on the topo section the sample vcf, in the middle the BWA-MEM2 BAM and on bottom the HaplotypeCaller output BAM):

chr1    164276519       .       G       A       156.50  PASS    AC=1;AF=0.500;AN=2;DP=1;ExcessHet=1.7609;FS=0;MQ=58.45;NEGATIVE_TRAIN_SITE;QD=38.82;SOR=1.981;VQSLOD=-1.284e+00;culprit=QD      GT:AD:DP:GQ:PL  0/1:1,0:1:24:56,0,24

The sample is a WGS, alinged with BWA-mem2, duplicate removal with picard MarkDuplicates, baserecalibration with GATK, clipping with clipBam, VariantCalling with HaplotypeCaller (GVCF output) and joint genotyping with other two WGS samples. Everything with default parameter (except not using softclipped bases in HC).

0

• 

  Gökalp Çelik

  • April 05, 2024 11:58

  Hi LucaB

  The region that you indicate is a highly repetitive region where variants get very little to no evidence due to the complexity. 

  For your actual inquiry we recommend you to check the documentation we provide for the explanation. 

  https://gatk.broadinstitute.org/hc/en-us/articles/360035532252-Allele-Depth-AD-is-lower-than-expected 

  I hope this helps. 

  0

  Comment actions Permalink

Please sign in to leave a comment.