The Supplementary Note lists ~140 covariates used, including blood cell counts and genotypes of blood count-associated SNPs. Supplementary Figure 1 shows the proportion of variance explained (R2) attributable to covariates and the effect of covariate control on heritability (h2) and variance explained by common (c2) and unique environment (e2), where these values were measured using a covariance (ACE) model that includes additive genetic, common or shared environment, and non-shared environment terms. Variance components were not constrained to be positive), so the model would be unbiased for h2, and to indicate whether genetic non-additive effects (dominance) may be present (by estimating c2 as negative). Covariate correction notably increased evidence for highly heritable transcripts, while no transcript was significant for c2 (Supplementary Figures 1b–e), in contrast to the MuTHER study.8
