Combining epidemiologic designs to model genetic risks for psychiatric disorders

点击翻译按钮获取中文摘要

Combining epidemiologic designs to model genetic risks for psychiatric disorders It is now widely recognized that genetic predispositions constitute major risk factors for psychiatric disorders such as schizophrenia (SZ) and biopolar disorder (BP). Yet we lag behind in epidemiologic methods for illuminating complex genetic architecture, or the ways in which mutational patterns at the genotypic level combine with environmental and other factors to produce specific behavioral phenotypes. One obstacle is the difficulty of combining information across different genetic epidemiologic study designs, for instance, exploiting both family studies and case-control studies in simultaneous analyses. As a result, it is difficult to construct models which adequately allow for the genetic and epidemiologic background in which given genes function. We have developed a powerful statistical approach (based on the “PPL” framework) and built a specialized computational platform (KELVIN) for modeling genetic architecture. We now propose to apply KELVIN to data contained in repositories established by the NIH, leveraging publicly available data sets in order to formulate more comprehensive models of the genetic architecture of SZ and BP. Specifically, we will make use of the large collection of multiplex pedigrees with microsatellite genome scan data assembled by the NIMH Human Genetics Initiative (HGI); and the extensive case-control samples with genome-wide SNP data assembled by the Genetic Association Information Network (GAIN). By applying our methods for combining information across genetic epidemiologic study designs to data already in the public domain, we can leverage earlier public investments to perform powerful modeling of genetic risk factors for major psychiatric disorders.