Novel population-genetic methods for localizing targets of natural selection in diverse human genomes

项目摘要。 PI在群体遗传学方面的研究计划侧重于基于聚结的群体遗传学推断。 从全基因组的历史，并确定适应和疾病的遗传基础，在多个生物学- 从突变到基因再到基因子网络。在基因组时代，计算和统计 方法对于识别人类中的候选适应性和疾病相关突变至关重要， 通过连锁研究绘制地图具有挑战性，而且费用高昂。最先进的方法，扫描全基因组， 选择或与表型状态相关的标记通常应用于来自一个同源基因的样品， neous祖先，依赖于任意阈值来解释结果，并在基因组规模上产生结果， 可能很难与生物机制联系起来（例如，分析连锁块或滑动基因组 windows）。因此，尽管美国国立卫生研究院和世界各地的生物库进行了巨大的投资， 由于来自不同个体的大规模基因组数据集，用于分析此类数据集的方法落后。 这个应用程序描述了一系列的项目，通过回答人类的三个基本问题， 群体遗传学：（1）平衡选择在人类适应中扮演什么角色？(2)在多大程度上 适应性进化与非适应性过程塑造了人类基因组？(3)基因在多大程度上 人类特征的结构因祖先而异？未来研究计划的总体战略借鉴了PI的 专业知识结合理论，贝叶斯推理，人口遗传学和统计遗传学，以产生新的 分析多种族基因组数据集的模式和进化过程的框架。的 本MIRA申请中描述的研究结果将为以下方面的相互作用提供新的见解： 选择和动态种群历史在产生人类遗传多样性，同时确定不同的 选择模式塑造人类表型和疾病。

PROJECT SUMMARY. The PI's research program in population genetics focuses on the coalescent-based inference of population his- tories from whole genomes, and the determination of genetic basis of adaptation and disease at multiple biolog- ical scales—from mutations to genes to gene subnetworks. In the genomic era, computational and statistical methods are essential for identifying candidate adaptive and disease-associated mutations in humans, in whom mapping via linkage studies is challenging and costly. State-of-the-art approaches that scan genome-wide for signatures of selection or association with phenotype state are routinely applied to samples from one homoge- neous ancestry, rely on arbitrary thresholds for interpreting results, and produce results at genomic scales that can be difficult to connect to biological mechanism (for example, analyzing linkage blocks or sliding genomic windows). Thus, despite the enormous investments made by the NIH and biobanks around the world to generate large-scale genomic datasets from diverse individuals, methods for analyzing such datasets are lagging behind. This application describes a series of projects motivated by answering three fundamental questions in human population genetics: (1) what role has balancing selection played in human adaptation? (2) to what extent has adaptive evolution versus non-adaptive processes shaped human genomes? (3) to what extent do the genetic architectures of human traits vary by ancestry? The overall strategy for future research plans draws on the PI's expertise in coalescent theory, Bayesian inference, population genetics, and statistical genetics to produce new frameworks for analyzing patterns in and evolutionary processes underlying multiethnic genomic datasets. The outcomes of the research described in this MIRA application will give new insight into the interaction between selection and dynamic population histories in generating human genetic diversity, while determining the different modes of selection shaping human phenotypes and diseases.