Population-Genetic Studies for Association Mapping

描述（由申请人提供）：关联研究为定位导致疾病易感性和表型变异的等位基因提供了强有力的方法。由于种群遗传过程在产生疾病及其因果等位基因之间的统计关联模式方面起着核心作用，因此，了解人类种群遗传历史及其关联后果对于制定绘制疾病易感等位基因图谱的方法非常重要。我们提出了四个具体目标，这些目标将增强和利用人类群体的理论和经验群体遗传学知识，以推进通过关联制图确定疾病易感位点的前景。这项工作将通过结合数学理论、计算机模拟和人类种群遗传数据分析来完成。首先，我们将扩展通过种群结构分析基因型和疾病之间虚假关联的方法，以适应临床或空间分布的种群。新方法将有可能在比目前可能的更广泛的情况下减少由人口结构或分层引起的假阳性关联的发生。其次，我们将开发人类进化的种群遗传模型，使用近似贝叶斯计算来解释世界各地不同种群的单倍型变异模式。第三，我们将为遗传关联的复制研究开发一个统计分析框架，该框架考虑到所有人类都是由共同祖先的后裔联系在一起的事实。第四，我们将比较从基因型计算的遗传关联统计和从估计的单倍型计算的遗传关联统计的特性，并将确定单倍型统计比不需要单倍型估计的方法提供更准确的关联信息的情况。这项工作将能够更准确地估计连锁不平衡，这在关联研究设计和分析中很重要。该项目的长期目标是最佳地利用人类变异和进化历史的知识来设计和分析关联图研究。我们的工作将利用全基因组微卫星和单核苷酸多态性数据，我们已经在世界范围内收集的人群。作为该项目的一部分，我们将开发新的统计方法，并在软件工具中实现它们，我们将使其公开可用。

DESCRIPTION (provided by applicant): Association studies provide a powerful approach for locating alleles that contribute to disease susceptibility and phenotypic variation. Since population-genetic processes play a central role in generating patterns of statistical association between diseases and their causal alleles, an understanding of human population- genetic history and its consequences for association is important for the development of methods to map disease susceptibility alleles. We propose four specific aims that will augment and capitalize on theoretical and empirical population genetics knowledge of human populations to advance the prospects for identifying disease susceptibility loci by association mapping. This work will be performed through a combination of mathematical theory, computer simulation, and analysis of human population-genetic data. First, we will extend methods for analysis of the production by population structure of spurious associations between genotypes and disease to accommodate clinal or spatially distributed populations. The new approaches will make it possible to reduce the occurrence of the false positive associations that arise from population structure or stratification in a broader set of scenarios than is currently possible. Second, we will develop population-genetic models of human evolution that use approximate Bayesian computation to account for patterns of haplotype variation among diverse worldwide populations. Third, we will develop a framework for statistical analysis of replication studies of genetic association that takes into account the fact that all humans are related by descent from shared ancestors. Fourth, we will compare properties of genetic association statistics computed from genotypes and from estimated haplotypes and will identify scenarios in which haplotype statistics provide more accurate association information than methods that do not require haplotype estimation. This work will enable more accurate estimation of the linkage disequilibrium important in association study design and analysis. The long-term goal of the project is to make optimal use of knowledge of human variation and evolutionary history for the design and analysis of association mapping studies. Our efforts will make use of genome-wide microsatellite and single-nucleotide polymorphism data that we have gathered in a worldwide collection of populations. As part of the project, we will be developing new statistical methods and implementing them in software tools that we will make publicly available.