The techniques currently available for gene identification remain controversial, a testimony to the difficulty in locating common disease genes. Genes contributing to psychiatric diseases with the greatest effect on public health, e.g., schizophrenia, bipolar, depression, as well as other complex nonpsychiatric diseases, still resist elucidation through genetic analysis. During the last grant period, we developed new computational methodologies for psychiatric diseases, by first determining how the analysis methods actually behave, then using that knowledge to create statistical genetic methods that we then tested and documented. Our productive research during the last grant period focused on understanding linkage analysis methods in common disease. We answered questions we had posed concerning: 1) The power of affecteds-only methods and lod scores; 2) The best family collection approaches; 3) Type I error in linkage analyses; 4) Quantifying heterogeneity in linkage analysis of complex diseases. Finally, 5) We applied our methods and insights to the analysis of real data, both psychiatric and other common diseases. In this renewal we focus on evaluating the statistical robustness of methods in the following areas: I. Linkage methods for understanding and detecting gene-gene interactions. Association analysis methods in presence of population stratification. Methods to detect imprinting. We use both analytical mathematical models and computer simulation as tools to accomplish these goals. Relevance: This work will contribute to public health by developing improved methods for identifying genes and understanding the genetics of common, complex diseases, including, but not limited to, psychiatric diseases. ? ?
Showing the most recent 10 out of 123 publications
