The project will provide computationally efficient statistical methods that explain genetic polymorphism affected by mutation, selection and genetic drift. Statistical analysis of this type of genetic data is complicated and computationally intensive. The new approach considered here combines methods from several mathematical disciplines. Techniques from numerical analysis, such as fast-Fourier transforms, make the proposed algorithms much faster than current methods. Stochastic approaches that efficiently simulate data provide a more reliable assessment of the methodology.
Many of the current methods for uncovering the genetic basis of complex diseases in humans aim to exploit the relationships between genes at loci close together on the same chromosome. These patterns depend crucially on the genetic variation at the loci involved. Consequently, there is considerable interest in understanding how these would be affected by selection. Additionally, this project will develop widely accessible, reliable software for analyzing genetic polymorphism in a way that will assess the impact of selection. Training will be provided to a graduate student who will be part of a newly formed interdisciplinary graduate program in Bioinformatics and Computational Biology (BCB) at the University of Idaho. This future scientist will leave the program with dramatically different training than the student's advisor. By being on the boundary of several disciplines, the student will be on the cutting edge of the future of computational biology.
