Mixed Model Approaches to Estimating Gene-by-Gene Interaction and Gene-by-Environment Interaction Based on Four Omics Data

GWAS can be conducted based on association of genotype to phenotype. Currently the mostly used genotype data is SNPs. We have developed mixed model approaches for estimating epistasis and environment interaction. The linear model is y = E + Q + QQ + QE + QQE + error. QTX mapping can be conducted by using SNP and gene methylation for quantitative trait SNP (QTS), mRNA and miRNA for quantitative trait transcript (QTT), enzyme and protein for quantitative trait protein (QTP), and alkaloids, terpenoids, glycosides, natural phenols for quantitative trait metabolite (QTM).  A software QTXNetwork has been developed by applying GPU parallel computation, so that it can efficiently analyze high through-put Omics genotypic data. Worked examples of human, mouse, and Drosophila are presented for reveling complicated genetics architecture of complex traits.