EQTL Mapping

Expression quantitative trait loci (eQTL) mapping concerns elucidating which transcripts orgroups of transcript are associated with which markers or sets of markers. This problem posessignificant challenges due to high dimensionality of both the gene expression and genomicmarker data. We propose a multivariate response regression approach with simultaneous variableselection and dimension reduction for the eQTL mapping problem. In our approach, transcriptswith similar expression are clustered into groups, and their expression profiles are viewed asmultivariate responses. Then, we employ our recently developed sparse partial least squaresregression methodology to select markers associated with each cluster of genes. We demonstrate,with extensive simulations, when and why this multivariate response approach gains power. Weillustrate that it compares competitively with other approaches for this problem and has a numberof significant advantages including the ability to handle highly correlated genotype data andcomputational efficiency. This framework bypasses the issue of multiple transcript-ormarker-specific analyses, thereby avoids potential elevation of Type-I error. Additionally, jointanalysis of multiple transcripts by the means of multivariate response regression leads to increasein power for detecting weak linkages.