After normalization of the data, we performed both cis- and trans-eQTL mapping. eQTLs were deemed cis-eQTLs, when the distance between the SNP chromosomal position and the probe midpoint was less than 250 kilobases (kb), while eQTLs with a distance greater than five megabases (mb) were defined as trans-eQTLs. Only SNPs with a minor allele-frequency (MAF) > 0.05 and a Hardy-Weinberg equilibrium p-value > 0.001 were included in the analyses. Since most cohorts had generated the gene expression data using the HT12v3 platform, we chose to only include probes that were present on this platform. We only tested SNP-probe pairs when the SNP passed quality control in at least three cohorts. Furthermore, in order to reduce issues with respect to computational time and multiple testing, we confined our transeQTL analysis to those SNPs present in the “Catalog of Published GWAS” (http://www.genome.gov/gwastudies/, accessed July 16th, 2011). We reasoned that for genes with strong cis-eQTL effects, the cis-eQTL effect may obscure the detectability of trans-eQTL. Therefore, we used linear regression to remove cis-eQTL effects prior to trans-eQTL mapping and observed a 12% increase
