Development of mapping by admixture linkage disequilibrium for understanding human complex genetic diseases

Mapping by Admixture Linkage Disequilibrium (MALD) is a theoretically powerful technique for mapping complex genetic diseases that is being developed for experimental application. The basic idea of MALD is to utilize the new linkage disequilibrium created in recently admixed populations to map disease susceptibility genes. This dissertation concerns the development of prerequisites in tools and knowledge needed for application of MALD to the African American and Mexican American populations. One practical requirement for MALD is a set of markers spanning the genome with large differences between the parental ethnicities. Toward this end, I developed a method to analyze genotyping results from pooled DNA samples, which can be used to rapidly determine marker allele frequencies in ethnic populations. This method performed better than previously published methods on pools of 50, 100 and 200 individuals. With the help of this method, over 1600 microsatellite, insertion/deletion, and single nucleotide polymorphism markers were screened for differences between European American, Amerindian, African, African American and Mexican American populations. In addition, many of these markers were examined in two African (Nigerian and Zimbabwean) or three Amerindian (Pima, Mayan, and Quechuan) tribes. Results demonstrate that markers with large inter-population and small intra-population differences can be readily identified for MALD studies in African American and Mexican American populations. MALD also requires admixed individuals to have genomes composed of large segments derived intact from each founding population. To determine if today's African American population meets this criteria, the extent of admixture linkage disequilibrium was examined in detail in the African American population using over 250 marker-pairs. Significant LD was observed over large genomic segments exceeding 10 cM and 15 Mb, although strong LD was not observed for markers separated by more than 10 kb in either founder population. Thus the large LD due to admixture theoretically predicted for the AA population was verified experimentally using attainable markers. The cumulative results of this dissertation demonstrate that MALD should be a feasible and powerful method for locating susceptibility loci in many complex genetic diseases.