Notice:
If you happen to see a question you know the answer to, please do chime in and help your fellow community members. We encourage our fourm members to be more involved, jump in and help out your fellow researchers with their questions. GATK forum is a community forum and helping each other with using GATK tools and research is the cornerstone of our success as a genomics research community.We appreciate your help!

Test-drive the GATK tools and Best Practices pipelines on Terra


Check out this blog post to learn how you can get started with GATK and try out the pipelines in preconfigured workspaces (with a user-friendly interface!) without having to install anything.

confirming my analysis strategy to work with sex chromosomes

Hello,

So, I wanted to confirm/discuss as I need to deal with sex chromosomes in my project. I have come across this post: https://gatkforums.broadinstitute.org/gatk/discussion/4639/x-chromosome-gentyping, but I wanted to be sure as the pipeline changed since 2014.

I am working on a non-human mammal which has reference genome from a female (so only X chromosome present for mapping). I have 81 samples out of which I have a mix of male and female samples.

For the first run, when I was not so much concerned about sex chromosomes, I ran everything (autosomes and sex chromosomes) using the default method (diploid for all samples and chromosomes). For female samples, it is fine. (Correct me if I am wrong). Since there are two copies of the X chromosomes, they are diploid for a variant call.

For male samples, I run them with ploidy =1 in haplotypecaller (gvcf), as males have just one X chromosome? What about them in genomicsDBimport and GeontypeGVCF?

Answers

  • SheilaSheila Broad InstituteMember, Broadie admin

    @prasundutta87
    Hi,

    That all sounds fine for females. I think for males you simply run the same workflow but use ploidy = 1 for the X chromosome. For the Y chromosome, I think the best thing you can do is either not call variants or use the X chromosome as a reference just so you can get some basic calls as comparison to some reference. Are the Y chromosome reads mapped to the X chromosome already?

    -Sheila

  • prasundutta87prasundutta87 EdinburghMember
    edited August 2018

    Hi Sheila,

    Mapping of all the samples (male and female has been done) and the variants have also been called using default Haplotypecaller parameters and joint genotyping has also been done to get the final set of variants. The only hitch was the male samples.

    I have already checked the no. of reads mapped to X chromosome for both male and female samples and found that male samples have approximately half no. of reads mapped to the X chromosome compared to the female samples. So, I believe that probably Y chromosome reads have not mapped to the X chromosome, but can't be sure, some reads may have mapped. Unfortunately, I don't have any reference genome of my non-model organism with a Y chromosome as the reference assembly is based on a female animal.

    Based on my already done analysis where I haven't changed the ploidy of the male samples, whichever male sample a variant on the X chromosome has been genotyped as a heterozygote, I have planned to remove it and leave X chromosome calls of female samples as they are.

    Also, I am concerned about the further steps, like genomicsDBimport and GenotypeGVCF. If at all I call variants on only the X chromosomes again by changing the ploidy for male samples only, will these steps remain the same?

    Post edited by prasundutta87 on
  • SheilaSheila Broad InstituteMember, Broadie admin

    @prasundutta87
    Hi,

    What happens if you change the ploidy to 1 for males on X chromosome? Can you post some example records for both het and hom var sites?

    Also, I am concerned about the further steps, like genomicsDBimport and GenotypeGVCF. If at all I call variants on only the X chromosomes again by changing the ploidy for male samples only, will these steps remain the same?

    This is not a problem, as both tools will infer ploidy from the input GVCFs.

    -Sheila

  • prasundutta87prasundutta87 EdinburghMember
    edited September 2018

    Hi,

    I haven't personally tested how haplotypecaller works by changing ploidy to 1, although, I am pasting some results below for two male samples for which ploidy was 2 (the default) in Haplotypecaller for X chromosome-

    I think, the GATK pipeline was able to correctly genotype variants (should be homozygous) present in the male X chromosome in its default setting (ploidy=2) in over 90 % of the data. Of course, the alignment done before by BWA-MEM and realignment in Haplotypecaller have also played a major role in this.

    Kindly let me know your views on this.

    -Prasun

    PS. Joint genotyping was done using 81 non-human samples (43 males and 38 females)

  • SheilaSheila Broad InstituteMember, Broadie admin

    @prasundutta87
    Hi Prasun,

    There are still a decent number of hets present. Can you post some example IGV screenshots of the BAM files at those sites (just a few will do).

    Thanks,
    Sheila

Sign In or Register to comment.