|
Abstract:
|
Complex diseases are characterized by multi-faceted biological responses to the environment, diet and often, treatment regimens. While single studies often lack the statistical power to identify casual variants of disease, meta-analysis provides the statistical framework to pool power across studies. Data harmonization across these studies, while often challenging, is required to apply meta-analysis techniques. The Mouse Phenome Database expertly curated studies are mapped to ontology terms providing the common semantic reference necessary to harmonize measurements of different aspects of the disease. A comprehensive genotype resource anchors the study populations. We leverage these resources to identify the causal genetic variants of complex diseases in the mouse. For each variant, conditional inference trees identify the conditions under which the effect exists, whether the effect is broad or specific to treatment, sex or other factors. Cross-species analysis is made possible through the variant mapping tool which links disease-causing variants in the mouse to human disease. This work is supported by NIH DA028420 and by The Jackson Laboratory, The Cube Initiative Program Fund.
|
