were also included as covariates. The effect of possible non-normality of the trait value on the Type I error rate of the linear regression analysis was tested by generating 10,000,000 random sets of genotypes for each of 3 minor allele frequencies 0.1, 0.25, 0.5. Regression of the trait on these genotype sets was performed, and Type I error rate estimated as the proportion of sets where the regression reached significance. No inflation of observed Type I error rate compared to nominal values was observed for any of the significance levels tested (ranging from 0.05 to 10-6). Thus, it was not considered necessary to perform a transformation on the factor scores.
