STATISTICAL METHODS AND APPLICATIONS IN HUMAN GENETICS

点击翻译按钮获取中文摘要

It is most likely that the genetic basis for the majority of common, familial diseases involves multiple genes, probably with interacting effects. Developing statistical- genetic study designs and methods of analysis to map and characterize genes for these disorders is the major challenge facing quantitative geneticists today. It is also critical to the success of the human genome project. This grant proposal is a competitive renewal requesting an additional five years of support to continue research in this important area. Specifically, it is proposed to (1) develop robust methods of linkage analysis in pedigrees, using Identity by descent for all pairs of affected relatives, based on marker information on all relatives and allowing for non-independence of pairs; (2) compare this robust method with conventional lod score analysis for different complex iodes of inheritance; different study designs (constellations of affected relatives) will also be contrasted for their power; (3) develop methods for simultaneous linkage analysis of hundreds of markers spanning the entire genome based on breeding experiments from animal models; (4) develop analogous methods for human linkage studies, based on simultaneous analysis of hundreds of marker loci spanning the genome; (5) examine the feasibility of marker association studies in recently admixed populations in a general genome search as compared with conventional linkage analysis; and (6) apply these methods to various data sets currently proposed for collection. These studies will combine both theoretical developments and computer simulations. The results from these analyses should have an enormous impact on methods for designing and analyzing genetic studies of common, complex familial diseases, and thus eventually to their full understanding.