Integrating Genomic and Genetic Data Across Species to Identify Gene Networks Contributing to Ethanol Behaviors and Alcohol Use Disorder


Mouse genomic studies on acute or chronic ethanol exposure models have been previously used to identify gene networks and cognate hub genes as targets for intervention in human alcohol use disorder (AUD). Similarly, multiple human genome wide association studies (GWAS) have been reported to identify genetic loci or genes conferring risk for AUD. However, neither of these approaches has, as yet, led to confirmed targets for development of future therapeutic approaches. This difficulty likely stems from the complexity of data produced by such high-throughput approaches and the small contribution that individual genes add to variance in ethanol behaviors. Our laboratory has therefore explored an approach for integrating genomic and genetic data across animal models and human GWAS results. Two approaches have been used to date. First, targeted mouse gene expression networks from acute ethanol exposure or chronic ethanol consumption models were assessed for over-representation of human GWAS signals. This verified a highly ethanol-responsive gene network as statistically associated with alcohol dependence at a network level. In a second approach, we directly integrated human GWAS data with mouse genomic data from acute or chronic ethanol exposure models in the BXD recombinant inbred panel. Using protein-protein interactions as a background matrix and a modified form of the dmGWAS program for network derivation, we identified several novel networks containing a high density of GWAS signals and ethanol-responsive genes. IGF signaling and extracellular matrix-related genes were particularly over-represented in two high scoring networks. Together, these studies hold promise for identifying new targets for experimental or therapeutic targeting, respectively, in animal models or human AUD. Supported by NIAAA grants U01AA016667, P50AA022537 AND R01AA020634 TO MFM.