We've moved!
This site is now read-only. You can find our new documentation site and support forum for posting questions here.
Be sure to read our welcome blog!

Identify haploid vs. diploid samples

beryljonesberyljones University of Illinois at Urbana-ChampaignMember

Hello! I have RNAseq data from 40 insects with a haplodiploid sex determination system. While for pupae and adults I know the sex of the individuals, I would like to determine the sex of the younger life stages. I believe there should be a way to use SNP variants to determine whether a particular library is from a haploid (male) or diploid (female) individual, but I don't know how the GATK tools work and what would be most appropriate. I would appreciate any advice!



  • SheilaSheila Broad InstituteMember, Broadie admin



    You can call all variants assuming they are all diploid. High quality variants in haploid individuals should all be homozygous; in diploids, there will also be heterozygous calls with good quality.

    Hope this helps.


  • beryljonesberyljones University of Illinois at Urbana-ChampaignMember
    edited April 2014

    Hi Sheila,

    Thanks for the response! Which variant calling tool would be best? Are there any designed for transcriptome rather than genome variant discovery?


  • KStammKStamm Member

    You're asking the GATK support forum; they are going to recommend HaplotypeCaller as the strongest variant caller possible on genomic samples. The older tool, UnifiedGenotyper should also work for these purposes, but none are designed for RNA-Seq. I know when I use files aligned by TopHat there are lots of false-positive variants hanging off exon boundaries, so you should carefully select exonic locations only for this purpose.

  • SheilaSheila Broad InstituteMember, Broadie admin



    We do have a new updated Best Practices that explains HaplotypeCaller's abilities to handle RNA-seq data. Please refer to it here: http://www.broadinstitute.org/gatk/guide/best-practices?bpm=RNAseq


  • charlesbaudocharlesbaudo MissouriMember

    Could you elaborate on how you assess those false-positive variants? The frequency of SNPs we saw in our RNA-seq SNP analysis were way higher than expected and I suspect that many will be false-positives. Thank you.

Sign In or Register to comment.