Dept. of Medical Informatics and Clinical Epidemiology
Oregon Health & Science University
Enormous data collection efforts and improvements in technology have made large-scale genome-wide association studies (GWAS) a promising approach to better understanding the genetics of common, complex diseases. However, the limited success of these studies so far suggests that genetic susceptibility may be due to a combined effect of multiple genetic variants (or interactions between variants), and that there may be a significant amount of genetic heterogeneity among those affected with complex diseases. It is clear that new data analysis methods are needed to address these hypotheses. Using data from the NIMH-sponsored Bipolar Genome Study, this project attempted to discover groups of SNPs that are jointly associated with the disease, thereby explaining a greater portion of disease susceptibility than can be achieved by examining SNPs individually. A machine-learning technique, known as a genetic algorithm, was used to search for these multi-locus associations, and was guided by a va
School of Medicine
Mooney, Michael A., "A computational method for detecting multi-locus associations :an application to bipolar disorder" (2011). Scholar Archive. 668.