Parametric linkage models, when carefully applied, can be invaluable in unraveling the complexities of interacting genes in complex diseases. The investigators will develop a faster likelihood engine using novel algorithms to extend the range of possible applications of parametric linkage analysis to include multilocus models to measure interaction between susceptibility genes and regressive models to be able to account for covariates, environmental factors, and familial correlations. Haplotype variation analysis is a very powerful approach to measure the variation of quantitative traits and to aid in their fine mapping. The investigators will develop zero-recombinant haplotype programs to efficiently carry out this analysis. The haplotype program will be particularly useful as array technologies begin to permit rapid genotyping at thousands of single nucleotide polymorphism markers. Simulation is an integral part in many non-parametric approaches and for testing the power of statistical models used in the study of complex diseases. A robust simulation engine based on the faster likelihood engine will be developed by the investigators. The algorithms will be implemented in VITESSE to develop an integrated package for the study of complex diseases using more powerful parametric models. The investigators will rigorously test and validate the computational and simulation engine and haplotyping programs to ensure correctness of the algorithms. They will also optimize the code to run efficiently on all platforms in general use by other researchers. The investigators will document and distribute the programs via the Internet to the worldwide scientific community.
