Homozygous VS Heterozygous

meharmehar Posts: 95Member

Hi,

I have two SNP calls as shown below:

chr1 86208198 . G C 903.25 PASS AC=2;AF=1.00;AN=2;BaseQRankSum=0.117;DP=33;Dels=0.00;FS=6.463;HRun=0;HaplotypeScore=8.5544;MQ=43.69;MQ0=0;MQRankSum=-0.506;QD=27.37;ReadPosRankSum=0.272;SB=-414.82 GT:AD:DP:GQ:PL 1/1:2,30:33:18.60:936,19,0

chr16 14895239 . C T 671.60 . AC=1;AF=0.50;AN=2;BaseQRankSum=-0.582;DP=33;Dels=0.00;FS=0.000;HRun=0;HaplotypeScore=1.5317;MQ=53.44;MQ0=0;MQRankSum=0.970;QD=20.35;ReadPosRankSum=-0.711;SB=-292.79 GT:AD:DP:GQ:PL 0/1:3,30:33:6.39:701,0,6

The first SNP is categorized as 1/1 and the second SNP as 0/1. For both the SNP's the variant allele ratio are 30/32=0.9375 and 30/33=0.909 which are approximately equal and above 0.9. On what criteria one SNP is determined as 0/1 and the other as 1/1?

As per my knowledge both the SNPs should be 1/1. Could anyone comment the reason for this discrepancy?

Tagged:

Answers

  • Geraldine_VdAuweraGeraldine_VdAuwera Posts: 9,962Administrator, Dev admin

    Have a look at this article, some of the points may be applicable: http://www.broadinstitute.org/gatk/guide/article?id=1235

    If not and you still think there is a problem, could you please post what version of GATK you are using and the command line?

    Geraldine Van der Auwera, PhD

  • meharmehar Posts: 95Member

    Hi,

    The second point seems to be applicable i.e. depth reported in the VCF is the unfiltered depth. But, here i have considered the filtered depth and im still unclear about the call.

    I am using GATK version 2.4-9 and the command line is:

    java -jar GenomeAnalysisTK.jar -glm SNP -R ref.fasta -T UnifiedGenotyper -I in.bam -o out.vcf.

  • Geraldine_VdAuweraGeraldine_VdAuwera Posts: 9,962Administrator, Dev admin

    Hmm. Could you run again on that interval using the latest version (2.6)? It would also help to see a screenshot of the pileup in IGV if possible.

    Geraldine Van der Auwera, PhD

  • meharmehar Posts: 95Member

    I have re-run it using GATK latest version and it called botht he SNPs as 1/1 and it looks fine. But what has been changed from the previous version to have a different call? Based on what criteria it calls as 1/1 or 0/1?

  • Geraldine_VdAuweraGeraldine_VdAuwera Posts: 9,962Administrator, Dev admin

    We have made some improvements to the calling model and fixed some bugs that led to incorrect calls. The most important changes to each version are listed here: http://www.broadinstitute.org/gatk/guide/version-history

    Geraldine Van der Auwera, PhD

Sign In or Register to comment.