Project 1: Incorporating Ethnic and Gender Disparities in Genomic Studies of Disease

项目摘要/摘要：我国医疗保健系统的负担大部分来自复杂的人力资源

Project Summary/Abstract: Most of the burden in our health care system comes from complex human diseases, whose onset and outcome are influenced by multiple genomic variants (e.g., cardiovascular disease, cancer, and diabetes). For the past decade, human geneticists have conducted genome-wide association studies, scanning for risk alleles associated with complex disease phenotypes. These studies have generally identified variants that confer relatively small increments in disease risk. We propose an alternative explanation for the unrealized promise of GWA studies in humans: rather than lacking the correct data with which to study medically-relevant traits, human genomics suffers from a lack of theoretical models that accurately characterize the population-level history of our species and the concomitant effects of natural selection. Overlooking population histories in the search for disease-associated variants leads to both spurious correlations between genotype and disease status, as well as the identification of disease-associated variants in one population that are not reproduced across multiple ancestral genomic backgrounds. Many common diseases vary in incidence across ethnicities and/or sexes, but association studies offer no framework to identify risk alleles for such diseases. The objective of this application is to incorporate the shared demographic history of human populations and models of variable dominance into genome-wide association studies. The central hypothesis is that human demographic history drives the incidence of common diseases across ethnicities and sexes. The aims of the proposal are to: 1) develop computational methods to identify risk alleles for diseases with disparities in incidence across ethnicities; 2) develop population-genetic methods to infer the fitness effects of mutations associated with diseases occurring differentially across sexes; and 3) apply newly developed methods to dbGAP association-study data from diseases with ethnic and sex-based disparities in incidence. The methods developed will be applicable to genome-wide, whole-genome and exome studies of disease association.