Genome Analysis Toolkit

Variant Discovery in High-Throughput Sequencing Data

GATK process banner

Need Help?

Search our documentation

Community Forum

Hi, How can we help?

Developed in the Data Sciences Platform at the Broad Institute, the toolkit offers a wide variety of tools with a primary focus on variant discovery and genotyping. Its powerful processing engine and high-performance computing features make it capable of taking on projects of any size. Learn more

Haplotypecaller missed variants in HLA genes

0

3 comments

  • Avatar
    danilovkiri

    Hi ensel

    I hope this veeeery old thread might happen to be extremely helpful in your case https://gatkforums.broadinstitute.org/gatk/discussion/5907/lower-gatks-haplotypecaller-threshold-for-allele-frequency-as-part-of-a-variant-calling-pipeline

    Generally, I advise you to just google your query prior to posting the question here. Googling "set low VAF for gatk haplotypecaller" provides you with the above-mentioned link.

    0
    Comment actions Permalink
  • Avatar
    ensel

    Thank you danilovkiri,

    and sorry for giving questions without googling.

    I got the idea how to make Haplotypecaller call low VAF, but is there any side effects when increasing ploidy?

    I am developing an analysis pipeline to genotype HLA genes for organ transplanting, and genotype accuracy is critical in clinical field.

    The NGS data are generated from PCR products of entire HLA genes, and the extreme biases of allele balance often occurs. 

    The VAF could be as low as 0.05 (5%), so the required ploidy would be about 20 which is extremely higher than 2. 

    Does 'ploidy' involves the expected VAF only, or could it affect others? 

    If high ploidy could result in incorrect variant calling (false positive/negative calls), what number of ploidy is tolerant to the side effects?

     

    0
    Comment actions Permalink
  • Avatar
    danilovkiri

    I guess the ploidy parameter is the last thing to mess up with due to the fact that it does affect other metrics. When referencing that old thread I mostly thought of Mutect2. It is a somatic variant caller and can handle extremely low VAFs in a correct manner, so I advise you to try it. 

    Additionally, the `--ploidy N` parameter results in N values in the GT FORMAT field (if ploidy is 10, then the GT might be something like 0/0/0/0/0/0/0/1/1/1), so it would require some intricate processing of such genotypes.

    There are multiple probabilistic HLA genotyping tools available (OptiType, for instance, https://github.com/FRED-2/OptiType), though I presume they would not give you the accurate you are looking for (OptiType has 4-digit accuracy). Besides, OptiType operates only with HLA Class I alleles. There are some other tools, maybe you will find something appropriate.

    0
    Comment actions Permalink

Please sign in to leave a comment.

Powered by Zendesk