Generalizing Haplotype Models for Phylogenetics

Carnegie Mellon University, Pittsburgh PA

Investigators

Abstract

This project will develop a combination of novel computational methods to improve inferences of human genomic history from data on genetic variability between individuals. Large amounts of data have recently been gathered on the small genetic differences that distinguish one person from another. Such information is valuable for inferring how the genome has changed over the course of human history and for understanding the specific biological processes that have shaped it over time. These basic scientific results in turn have relevance to many practical problems in identifying the functions of genes, facilitating future genetic studies, and finding and mitigating risk factors for disease. Making optimal use of this information will, however, require new advances in mathematical models of how genetic variations are produced and in the computer algorithms needed to use those models to analyze real sequence data. The proposed work is based on a strategy to overcome limitations of the data and current computational capabilities by combining theoretical tools for haplotype structure identification, which finds segments of DNA that have been largely conserved over long periods of human history, with tools for phylogenetics, which infers how different individuals or populations are related to one another. Executing this strategy will involve a multidisciplinary effort, bringing together techniques from population genetics, computer science, mathematics, and operations research. In addition to its direct scientific outputs, the project will develop human resources by creating new course material and student research projects to help prepare future scientists to work at the intersection of these fields.