Diverse implications of population structure for identifying human disease genes

描述(申请人提供)：识别疾病基因是有效治疗人类疾病的重要一步。识别人类疾病基因的努力既被种群结构的存在所诅咒，也被种群结构的存在所祝福，这可能导致在一些疾病图谱方法中错误的疾病关联，但实际上推动了其他疾病的存在。这项研究建议旨在减轻这些诅咒，并利用这些祝福。第一个目标是改进现有的方法，以纠正全基因组单倍型关联研究的人口分层，即病例和对照之间的系统祖先差异，这是虚假疾病关联的一个重要来源。第二个目标是将全基因组祖先图谱应用于拉丁裔人口；重点将放在II型糖尿病上，这种糖尿病在美洲原住民中特别普遍，因此可能会导致拉美人中患有II型糖尿病的拉丁裔美国原住民祖先的比例增加。第三个目标是开发一种全基因组自然选择图谱的方法，该方法利用样本种群之间的差异来识别最近自然选择下的基因，这些基因与疾病有很强的相关性。

DESCRIPTION (provided by applicant): Identifying disease genes is an important step towards effective treatment of human disease. Efforts to identify human disease genes are both cursed and blessed by the existence of population structure, which can lead to false disease associations in some disease mapping approaches but actually drives the existence of others. This research proposal aims to alleviate these curses and harness these blessings. The first aim is to improve upon existing methods to correct whole-genome haplotype association studies for population stratification, i.e. systematic ancestry differences between cases and controls, which is a significant source of false disease associations. The second aim is to apply whole-genome ancestry association mapping to Latino populations; the focus will be on type II diabetes, which is particularly prevalent in Native Americans and may thus induce an elevated proportion of Native American ancestry in Latinos with type II diabetes near a disease gene. The third aim is to develop a method for whole-genome natural selection mapping which harnesses the differentiation between sampled populations to identify genes under recent natural selection, which are strong candidates for relevance to disease.