An extensive number of studies have characterized the level and patterns of regulatory variation and eQTLs over the last decade. eQTLs have been studied in lymphoblastoid cell lines (LCLs) from the HapMap populations [11], [12], [13], [14], [15], in single tissues such as fat, osteoblasts and brain (cortex) [16], [17], [18], as well as across multiple tissues and cell types of the same individual to enable direct comparison of differential cis-regulatory effects [19], [20]. Collectively, these studies have demonstrated an abundance of eQTLs in all tissues, and have revealed a substantial tissue- and cell-type specific component of eQTLs. While these studies have partly addressed the degree of population differentiation among human populations for eQTLs [15], [21], they have been limited to only several well-defined populations, and in particular have not contrasted geographically proximate populations. In this study we describe the first analysis of eQTL differentiation among eight human population samples, including three populations from Africa and four admixed populations, using genome-wide expression data for 726 individuals and dense genotyping of over 1.2 million common SNPs; this comprises the most comprehensive dataset and analysis used for this purpose to date.
