Measuring and using genetic ancestry information in genome-wide admixture mapping and association mapping of complex diseases

There exists much variation in genetic ancestry within and between ethnic groups, which causes substantial population stratification to be present not only in recently admixed populations like African Americans but also in generally assumed homogeneous populations like European Americans. In Chapter One I reviewed the recent studies of measuring and using genetic ancestry in human complex disease studies. Genetics variations constitute an important basis for Admixture Mapping. Many complex diseases show population specific prevalence that could be due to the differences of particular disease-susceptible genes among founding populations of different ancestry. Statistical methods can be applied to infer the locus ancestry along the chromosome in admixed individuals and tests for the association of the locus ancestry with the disease in admixed population, so called admixture mapping. Admixture mapping requires a genome-wide panel of relatively evenly spaced markers that can distinguish the locus ancestral origins in admixed individuals. In Chapter Two and Chapter Three I introduced our defined genome-wide Single-Nucleotide-Polymorphism panels that can extract ancestry information mostly with the least markers for African American and Mexican American admixed populations. On the other hand, a consequence of population stratification is the potential for false allelic associations and thus the inconsistent reports across genome-wide association studies. Statistical methods can be applied to discern and correct for the individual ancestry differences using Genome-wide association panel. In Chapter Four I introduced our findings of the European substructures, which have significant genetic variation along the north to south and west to east geographic axis. One of our recent report showed that after accounting for genetic ancestry difference, some locus are no long associated to Rheumatoid Arthritis but they appeared as very strong candidates without accounting for the substructure. In Chapter Five I introduced our findings of the East Asian substructures. Our analysis showed that there exist genetic variations both between different East Asian groups and within the Han Chinese population. In Chapter Six I reviewed the current available methods and importance of accounting for ancestry in genome-wide association studies. In Chapter Seven, I discussed some implications and future research directions.