Making use of phased haplotypes in rare variant burden testing

Question

[vlaufer]

Hello all,

I have a GWAS cohort and a WGS cohort that is a subset of the GWAS cohort.

I have identified all the lead SNPs from the GWAS data (larger n) and have defined linkage disequilibrium around each lead SNP in the study.

Further, I have phased the WGS genomes. As such, I can determine the variants that are in phase, or not in phase, with the lead SNPs in the study.

Right now, it is my intention to do this by parsing the VCF files, then reformatting into SKAT-O format. However, my question is, are there existing workflows that port phased genomes directly into any rare variant association tool, or, is one or more of these steps DIY?

For clarity, what I mean by rare variant association tools is roughly summarized by the Lee / Lin review paper in AJHG

Answer 1

[brentp]

I wrote a script to do exactly this:

It parses the VCF in python and sends the haplotype blocks to R using pyper. It can be a start for the parsing/creating haplotypes or you can use it as-is to get p-values from SKAT.

There are a lot of methods for testing haplotypes, but I think this poster does a nice job comparing.

[vlaufer]

Brent -

Thank you very much for sharing your code - very kind of you. I will accept this answer after a little time elapses, I do not want to discourage other answers.

Note that you have the close parenthesis as part of the hyperlink, causing the link to fail without modification.