Methods for genetic linkage analysis in the presence of heterogeneity

Genetic heterogeneity is one of the most significant obstacles to identifying the genetic basis for many common human diseases. Heterogeneity is the term used for genetic systems in which numerous genes each make a small contribution to the overall heritability of a disease. Linkage analysis has been used successfully for several decades to map disease susceptibility genes, but it lacks power to identify susceptibility genes in heterogeneous systems. The purpose of this research is to improve current methods and develop new methods for linkage analysis in the presence of genetic heterogeneity. Competency is established in conventional and emerging methodology, new methods are developed, and the newly developed methods are tested in real study data. Prostate cancer (PCa), a prime example of the problems that heterogeneity creates for genetic epidemiologists, is used as a model system throughout the research. Studying alternate PCa phenotype definitions or PCa subtypes may improve our knowledge of the disease. Chapter 2 describes a conventional linkage analysis for aggressive PCa subtypes, the results of which confirm two previously reported PCa aggressiveness loci. Chapter 3 presents proof of concept that phenotypes based on gene expression profiles from microarray data may be useful for identifying genes associated with risk of PCa development via linkage analysis. Chapters 4 and 5 describe the development and application of the innovative sumLINK statistic, which identifies genetic regions of extreme consistency across pedigrees without regard to negative evidence from unlinked or uninformative pedigrees. Significance of the sumLINK statistic and the complimentary sumLOD statistic is determined empirically by an innovative permutation procedure that randomizes linkage information across pedigrees. Simulation testing shows that this method is reliable and powerful for finding genes in heterogeneous systems. The utility of the sumLINK method is demonstrated with exciting results using data from the International Consortium for Prostate Cancer Genetics for aggressive and general prostate cancer. The sumLINK procedure fills an important informatics role by facilitating secure interinstitutional data sharing and collaborative research. The sumLINK method is a powerful tool for combating the obstacles presented by heterogeneity, and will improve our knowledge of the genetic epidemiology of many common, complex diseases.