We obtained whole blood RNA-Seq38 and genome-wide genotype data for 922 individuals from the Depression Genes and Networks cohort16, all of European ancestry. For our analyses, we used the HCP (hidden covariates with prior) normalized gene-level expression data used for the trans-eQTL analysis in Battle et al.16 and downloaded from the NIMH repository. Approximately 650K SNPs (minor allele frequency [MAF] > 0.05, Hardy-Weinberg Equilibrium [P > 0.05], non-ambiguous strand [no A/T or C/G SNPs]) comprised the input set of SNPs for imputation, which was performed on the University of Michigan Imputation-Server39,40 with the following parameters: 1000G Phase 1 v3 ShapeIt2 (no singletons) reference panel, SHAPEIT phasing, and EUR population. Non-ambiguous strand SNPs with MAF > 0.05, imputation R2 > 0.8 were retained for subsequent analysis. To reduce computational burden in the application to WTCCC, we used models developed on the HapMap Phase II subset of SNPs.
