Stochastic Processes in Population Genetics

点击翻译按钮获取中文摘要

DESCRIPTION (provided by applicant): The main aim of this proposal is to bring together various aspects of work which the PI has been involved with over many years in the three areas of evolution, linkage analysis and bioinformatics. to address questions concerning diseases in humans, evolution, and the analysis of data from the Human Genome Project. Evolutionary population genetics theory has in large part considered one gene locus and random-mating populations. Both of these restrictions are inappropriate for the analysis of data from the human genome. The human population does not mate at random, and whole genome data implies that an analysis considering arbitrary collections of loci is required. Part of the proposed research involves construction of theory overcoming these restrictions. This will be linked with work currently being performed by the PI in bioinformatics, in particular on theory associated with the BLAST procedure and evolutionary phylogenetic models. A major aspect of the proposed research will concern new methods for finding disease genes, using whole-genome data. For example, whole-genome sharing data is now available for collections of descendants of an affected individual. The analysis of these data requires new statistical methods, generalizing the genomic mismatch scanning methods already completed. Other aims concern further research on specific topics. As one example, questions from users continually arise concerning the transmission/disequilibrium test (TDT), introduced by the PI eight years ago and now widely used for linkage and association studies concerning diseases in humans. These often require new theory, as does the general development of TDT theory.