Does the UG call variants in pseudoautosomal regions correctly?

evakoe

Dear GATK team,

when I use the UnifiedGenotyper to call variants in chrX and/or chrY with ploidy=2, does the program call the variants and genotypes in the pseudoautosomal regions PAR1 and PAR2 correctly? More specifically, does the UG acknowledge that these regions are homologous on the X and Y chromosome, even though they differ in chromosomal position? If the UG does not take care of this by default, is there a way I can handle this? I read point 6 of this page, but I could not find the answer there: http://www.broadinstitute.org/gatk/guide/article?id=1237

Thanks alot Eva

Kurt

If i'm not mistaken, if using the human reference genome in the gatk resource bundle, then the PAR regions are not present on chrY, just X (I don't have access to any data at the moment, but you should be able to check the sequence dictionary start/stop in your reference genome file for chrY to see if they exclude the PAR). I'd imagine if both were present bwa would ding mappings to PAR because a read could map to both X and Y. So I would think that you really don't have to do anything special if using that reference other than what @evakoe is asking.

Geraldine_VdAuwera

Hi Eva, have you looked at this presentation?

https://www.dropbox.com/sh/55nfktmn7lgai98/z1tCcu9ngA/UG on sex chromosomes v1.pptx.pdf

evakoe

Yes, if I understand correctly, the PAR regions of the Y are mapped to the corresponding positions in the X chromosome. So I want to call SNPs in the PAR regions in a male sample, I have to call in the regions chrX:60001-2699520 and chrX:154931044-155260560. Is that right? Thanks. Eva

Geraldine_VdAuwera

Yes, that's correct, but I believe you'll also have to call over the analogous regions in the Y chromosome, and you'll have to somehow manually merge variants corresponding to the same locations to get true het states if that's what you're after. We don't treat those locations in any special way - they're just different contigs to us and all merging/analysis of the PAR1 and PAR2 regions will have to be done by users or downstream tools.

Geraldine_VdAuwera

Thanks for the clarification, @Kurt, that sounds right.

evakoe

Great thanks alot Kurt and Geraldine. This helped me much. Eva