Gene expression in the mouse eye: an online resource for genetics using 103 strains of mice

PURPOSE: Individual differences in patterns of gene expression account for much of the diversity of ocular phenotypes and variation in disease risk. We examined the causes of expression differences, and in their linkage to sequence variants, functional differences, and ocular pathophysiology. METHODS: mRNAs from young adult eyes were hybridized to oligomer microarrays (Affymetrix M430v2). Data were embedded in GeneNetwork with millions of single nucleotide polymorphisms, custom array annotation, and information on complementary cellular, functional, and behavioral traits. The data include male and female samples from 28 common strains, 68 BXD recombinant inbred lines, as well as several mutants and knockouts. RESULTS: We provide a fully integrated resource to map, graph, analyze, and test causes and correlations of differences in gene expression in the eye. Covariance in mRNA expression can be used to infer gene function, extract signatures for different cells or tissues, to define molecular networks, and to map quantitative trait loci that produce expression differences. These data can also be used to connect disease phenotypes with sequence variants. We demonstrate that variation in rhodopsin expression efficiently predicts candidate genes for eight uncloned retinal diseases, including WDR17 for the human RP29 locus. CONCLUSIONS: The high level of strain variation in gene expression is a powerful tool that can be used to explore and test molecular networks underlying variation in structure, function, and disease susceptibility. The integration of these data into GeneNetwork provides users with a workbench to test linkages between sequence differences and eye structure and function.