GATK-HaplotypeCaller results generate records with read depth of zero?

Follow

xie186

• June 25, 2020 14:08

a) GATK version used: GATK/4.1.4.1
b) Exact GATK commands used: 

gatk HaplotypeCaller -R Pfalciparum.genome.fasta -I recal_realn.bam -ploidy 2 -ERC GVCF --standard-min-confidence-threshold-for-calling 30.0 -O HapoCaller.raw.g.vcf

I found that "DP" of some of the positions are ZEROs (e.g. the one shown below).

Pf3D7_07_v3 777185 . T *,A,<NON_REF> 38.86 . DP=6;ExcessHet=3.0103;MLEAC=0,2,0;MLEAF=0.00,1.00,0.00;RAW_MQandDP=21600,6 GT:AD:DP:GQ:PL:SB 2/2:0,0,0,0:0:9:162,22,109,13,9,0,102,75,20,136:0,0,0,0

Could you please let me know why DP is ZERO in "2/2:0,0,0,0:0:9:162,22,109,13,9,0,102,75,20,136:0,0,0,0"? Thanks. 

 

 

 

1

• 

  Bhanu Gandham

  • June 30, 2020 00:03
  • Edited

  Hi xie186

  There is often confusion between "INFO" field DP, "FORMAT" filed DP and "FORMAT" field AD values. Here are some definitions that might help bring some clarity.

  INFO DP - includes all reads that overlap a variant after local realignment, that passes the quality filters
  FORMAT DP - includes only informative reads for that sample
  FORMAT AD - only informative reads for each allele; sum of all AD = DP(unless the alleles exceed the max alt alleles limit)

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

   

  In this case you see DP = 0 since there are no informative reads for that variant.

  0

  Comment actions Permalink
• 

  xie186

  • June 30, 2020 12:57

  Thank you for your explanation, @Bhanu Gandham.

  I really appreciate it.

  0

  Comment actions Permalink

Please sign in to leave a comment.