SNPxGE2: a database for human 3-way SNP-expression associations

Abstract

Recently, gene co-expression relationships have been found to be often conditional and dynamic. Besides, many studies have suggested that single nucleotide polymorphisms (SNPs) have impacts on gene expression variations in human populations. The SNPxGE2 database contains the computationally predicted human 3-way SNP-expression associations, that is, the differential co-expression between 2 genes is associated with a genotype/SNP. This data was generated from a large scale association study that was based on the HapMap data, which covered 269 individuals from 4 human populations, 701,202 SNPs and 15,000 gene expression profiles. Two models of 3-way SNP-expression associations were considered: gap/substitution and on/off. The implementation was carried out using 64 Linux cluster nodes in ~30 days and assessed SNP-expression combinations. The results, including 4,713 on/off associations and 36,170 gap/substitution associations at a p-value cutoff of 0.001, can be queried in the SNPxGE2 database via either gene name or reference SNP ID. For each reported association, a detailed information page is provided. The SNPxGE2 database can be freely accessed at http://tunisia.ads.uga.edu/SNPxGE2/index.php.