Modeling the Nuclear Families Data in Human Genetics

Abstract

Family-based association have shown great promise in the recent and future studies of detecting genetic factors of complex disorders in human diseases. Among the many methods proposed during the past decade, the transmission/ disequilibrium test (TDT) of Spielman et al. (1993) and its extensions stand out as important tools for analyzing nuclear families data. Generally it is agreed that, for the detection of loci for complex disorders the TDT is a more powerful tool than linkage methods such as allele sharing. The TDT is so easy to compute and requires so few assumptions to carry out a valid analysis that it is easy to forget its limitations, especially when dealing with a complex trait whose responsible genes or candidate genes are not identified. From statistical point of view, it is desirable to have a general theory and a flexible framework which provide a comprehensive basis for the analysis of family-based data. We shall introduce a realistic model in this talk. We model the marginal distribution of allelic transmissions from one heterozygous parent to his/her affected children. Since the model is built on the parental genotypes, the confounding effects and influences of all nuisance parameters such as allele frequencies, modes of inheritance, penetrance, mating types and population stratifications are automatically eliminated. If time permits, further aspects such as advantages and limitations, prospective research and computational issues will be discussed.