|
Post by weihuameng on Feb 17, 2016 15:18:02 GMT
Dear: Hi. Is the following GWAS design correct using GCTA?
stage1: 2000 cases vs 2000 controls, I got a smallest p value =10-6 using MLMA
stage 2:another 2000 cases vs 2000 controls, I got a smallest p value =10-7 (same SNP) using MLMA
HOWEVER, the samples used in the second stage is related with samples used in the stage 1 or related with other samples in the stage 2 as well.
I tried to combine stage 1 and 2, 4000 cases and 4000 controls , I got a smallest p value(same SNP) of 10-8.
since MLMA will consider relatedness, i assume it is a fine GWAS design (IN PRINCIPAL)?
Thank you. Weihua
|
|