Integration of Diverse Statistical Evidence of GeneTrait Association in Systems Biology Studies

The rapid advancement of high-throughput genomic assay technologies has generated large amounts of diverse genomic data in disparate human populations and diseases. These data provide a unique opportunity for biomedical investigators to systematically study multifaceted aspects of genes' involvement in the biological processes underlying important traits from the systems biology perspective. An important component in such a study is the inference that integrates diverse lines of statistical evidence for genetrait association from the observed trait values and the massive numbers of measured genomic features. A novel integrated statistical analysis procedure is developed in this paper and is illustrated by an application in studying childhood leukemia.