Haplotype-based GWAS

[anujbio...@gmail.com]

Hi,

I have a vcf file of 200 samples and I am planing to contstruct  haplotype blocks based R2 for GWAS run. Could you please help to contruct haplotype blocks and haplotype blocks file to convert into  bed file so I can use the bed file for GWAS?

 Is plink 2.0 can do  for both jobs (contruction of haplotype bloacks and hap blocks file to bed file)?

Thanks

[Christopher Chang]

0. I doubt 200 samples can provide enough statistical power to make this analysis worthwhile, unless the effect size is huge.

1. You can use plink 1.9's --blocks flag to estimate haplotype blocks.  (Note that this flag has not been implemented in plink 2.0, and I haven't yet decided whether it will be.)

2. No, you will need to use another software package to re-encode the VCF in terms of haplotype blocks instead of single variants.  However, once you do that, you can use plink 2.0 --glm for the final GWAS step; see https://groups.google.com/g/plink2-users/c/xqUw29voUR0/m/U9n6AF6FAwAJ for some discussion of this.