The main focus of this project is to develop statistical methods for mapping genes affecting polygenic and multi factorial traits (referred to below as Quantitative Loci or QTL) in large, complex pedigrees. The project has three interrelated objectives: (i) We are expanding our QTL mapping software to the analysis of multi-generational pedigrees, to choice among models with different numbers of QTL polymorphism, and to discontinuous but polygenic (disease) phenotypes; (ii) we are comparing and providing tolls for model checking and choice for Bayesian analysis of Generalized Linear Mixed Models (GLMMs) in general and QTL models in particular, as well as investigating alternative sampling algorithms necessary to implement these methods, and (iii) We are investigating strategies for the fine mapping of QTL to enable positional cloning of important QTL and to increase the efficiency of marker-assisted selection in livestock. The availability of mderate resolution maps in livestock an high resolution maps in humans has created a substantial demand for statistical methods and tools which make optimum use of the available data and are capable of handling large, complex pedigrees, unknown number of QTL, and unknown nature of genetic variation at QTL. Current methods do not yet have all of these capabilities. However, the importance of methods and software extends far beyond livestock. They are critical to understanding and mechanisms and evolution of polygenic and multifactorial traits, designing more effective breeding regimes for domesticated plants and animals, and understanding and treating inherited human diseases. This work is supported by the Computational Biology Activity and the Program Population Biology (Bio).
