Haplotype Caller: too many alternative alleles found?

Hello gatk team,
I am running HaplotypeCaller on 5 files of genomic alignments together at once.
Despite the fact that I did a InDel realignment (Indel Realigner gatk) before running these files with HaplotypeCaller, I get many warnings during the process that say "too many alternative alleles found", with sometimes 10, 12 or 13 alternative alleles found.
Is that normal? or is there a step that I could have done improperly?
Thank you for your help :)
Marvin

Tagged:

Best Answer

  • SheilaSheila Broad InstituteMember, Broadie, Moderator
    Accepted Answer

    @mac
    Hi Marvin,

    It is possible to have many alternate alleles at a site. They usually occur in very messy regions or at sites where the samples are very divergent. We usually set --max_alternate_alleles 6. If you are running on 5 samples at once or on non-diploid samples, you may consider setting the default to a number higher than 6.

    You can also take a look at the sites that are in the WARN statements and check in IGV if they are messy sites. Those sites may not be worth increasing the max_alternate_alleles for, as the tool will not be able to produce confident calls at those sites.

    -Sheila

Answers

  • SheilaSheila Broad InstituteMember, Broadie, Moderator
    Accepted Answer

    @mac
    Hi Marvin,

    It is possible to have many alternate alleles at a site. They usually occur in very messy regions or at sites where the samples are very divergent. We usually set --max_alternate_alleles 6. If you are running on 5 samples at once or on non-diploid samples, you may consider setting the default to a number higher than 6.

    You can also take a look at the sites that are in the WARN statements and check in IGV if they are messy sites. Those sites may not be worth increasing the max_alternate_alleles for, as the tool will not be able to produce confident calls at those sites.

    -Sheila

Sign In or Register to comment.