Scoring proceeded in a series of steps. First, the phenotype was residualized using a linear regression on covariates of sex, generational status (parent or child), age, year of birth, and the first 10 principal components produced by EIGENSTRAT. A GWAS was then conducted on the residualized phenotype, producing a univariate regression weight for the minor allele count for each SNP. Minor allele counts for each SNP were then multiplied by their corresponding regression weight and summed to form a single score for each participant in the sample. This sumscore was then validated by correlating it with the residualized phenotype. Squaring the correlation gives the variance in the phenotype accounted for by the SNP score.
