Genetic Risk Score Analysis for POAG

We calculated 127-variant GRS for HIS and EUR Veterans in the MVP. GRS were either unweighted or weighted by published cross-ancestry effect estimates (14 (link)) as shown in Equations 1 and 2, respectively. Risk alleles were defined by having odds ratios greater than 1 in the cross-ancestry analysis (14 (link)).
$$GR{S}_{unweighted(i)}={\sum }_{j=1}^k{M}_{ij}$$
where M = risk allele dosage, i = individual, k = 127 variants
$${GRS}_{weighted(i)}={\sum }_{j=1}^k{\beta }_j{M}_{ij}$$
where M = risk allele dosage, i = individual, k = 127 variants, β=log(odds ratio)
We tested for association between the GRS and POAG via logistic regression-based analyses using unadjusted models as well as models adjusting for age, sex, and 10 sample-specific principal components (PCs).