ljc
New Member
Posts: 4
|
Post by ljc on May 19, 2015 8:04:18 GMT
Hi
I was wondering how X chr SNPs (chr 23) are handled when running --mlma? Do males and females need to be run separately? I am using bestguess genotypes from 1000 genomes imputed data.
Thanks
Laura
|
|
|
Post by Zhihong Zhu on May 19, 2015 13:04:47 GMT
Hi Laura, I think males and females can be run jointly. But I'm not sure about the coded genotypes of males, 0/1 or 0/2, in terms of dosage compensation. Maybe you could try running association by R using the residuals conditioned on GRM as phenotype, "--reml-pred-rand" ( more detailed information can be found at www.complextraitgenomics.com/software/gcta/reml.html). Cheers, Zhihong
|
|