Genomic Predictors of Arteriosclerosis in Hypertensives

描述（由申请人提供）： 动脉炎（即，动脉粥样硬化和小动脉硬化）导致靶器官损伤和临床后遗症，如心脏病发作和衰竭、中风和痴呆、慢性肾病和跛行。由于高血压的致病途径多种多样，大量基因的变异可能影响高血压个体对靶器官损害的易感性。动脉病遗传流行病学网络（GENOA）于1995年启动，旨在通过同胞关系连锁分析研究高血压及其靶器官并发症的遗传学。作为正在进行的GENOA研究的一部分，收集了动脉粥样硬化的定量指标，包括冠状动脉钙化的动脉粥样硬化/大血管指标、主动脉根直径、踝肱BP指数和缺血性脑损伤（皮质下白色高信号、脑萎缩和心室容积）和慢性肾脏疾病（血清肌酐和白蛋白尿）的小动脉粥样硬化/微血管指标。在本申请中，我们建议对这些动脉硬化表型指标进行全基因组关联研究，利用来自GENOA队列的1418名非洲裔美国人和1095名非西班牙裔白人的500，000个单核苷酸多态性。具体目标如下：目标1：使用500，000个SNP，在1418名非裔美国人和1095名非西班牙裔白色GENOA参与者中鉴定含有用于定量测量动脉粥样硬化的疾病易感性位点证据的基因组区域。我们将使用最先进的单变量和多变量方法对这两个样本进行全基因组关联分析，以确定性状特异性位点，以及对动脉硬化的多个遗传相关指标具有多效性效应的遗传位点。利用我们数据的高血压同胞结构，我们将创建两个重复的高血压患者组（一个同胞在无关组1中，第二个同胞在无关组2中），以进行重复的全基因组关联。SNP关联将针对多重测试进行调整，使用交叉验证评估其预测效用，并比较重复证据以减少假阳性。目标二：为了研究背景依赖性遗传效应对这些队列中靶器官疾病分布的作用，我们将检验SNP-协变量相互作用（年龄、性别、吸烟、肥胖、酒精、体育活动、教育、血压药物治疗）和SNP-SNP相互作用对目标1中研究的动脉粥样硬化相同定量指标的影响。SNP关联将针对多重测试进行调整，使用交叉验证评估其预测效用，并比较重复证据以减少假阳性。我们还将使用监督模式识别方法（分类和回归树，随机森林，基因组状态识别方法）来构建这些临床表型的预测模型，以识别风险个体。 (End摘要）

DESCRIPTION (provided by applicant): Arteriosclerosis (i.e., atherosclerosis and arteriolosclerosis) of the cardiac, cerebral, renal, and peripheral arteries leads to target organ damage and clinical sequelae such as heart attack and failure, stroke and dementia, chronic kidney disease, and claudication. Because of multiple etiologic pathways, variations in a large number of genes are likely to influence susceptibility to target organ damage in hypertensive individuals. The Genetic Epidemiology Network of Arteriopathy (GENOA) was initiated in 1995 to study the genetics of hypertension and its target organ complications using linkage analysis in sibships. As part of the ongoing GENOA study, quantitative measures of Arteriosclerosis including both atherosclerotic/ macrovascular measures of coronary artery calcification, aortic root diameter, ankle-brachial BP index, and arteriolosclerotic /microvascular measures of ischemic brain injury (subcortical white matter hyperintensities, brain atrophy, and ventricular volume) and chronic kidney disease (serum creatinine and albuminuria) have been collected. In this application, we propose to conduct a genome-wide association study of these phenotypic measures of arteriosclerosis, utilizing 500,000 single nucleotide polymorphisms measured on 1418 African-Americans and 1095 non-Hispanic Whites from the GENOA cohort. With the following specific aims: Aim 1: To identify genomic regions containing evidence of disease susceptibility loci for quantitative measures of Arteriosclerosis in 1418 African-American and 1095 non-Hispanic white GENOA participants using 500,000 SNPs. We will perform genome-wide association analysis in these two samples using state-of-the-art univariate and multivariate approaches to identify trait-specific loci, as well as genetic loci with pleiotropic effects on multiple genetically correlated measures of arteriosclerosis. Capitalizing on the hypertensive sibpair structure of our data, we will create two replicate sets of hypertensives (one sib in unrelated Set 1, second sib in unrelated Set 2) to perform replicate genome-wide associations. SNP associations will be adjusted for multiple testing, assessed for their predictive utility using cross-validation, and compared for replicate evidence to reduce false positives. Aim 2: To investigate the role of context dependent genetic effects on the distribution of target organ disease in these cohorts, we will test for effects of SNP-covariate interactions (age, sex, smoking, obesity, alcohol, physical activity, education, blood pressure drug treatment) and SNP-SNP interactions on the same quantitative measures of Arteriosclerosis investigated in Aim 1. SNP associations will be adjusted for multiple testing, assessed for their predictive utility using cross-validation, and compared for replicate evidence to reduce false positives. We will also use supervised pattern recognition methods (classification and regression trees, random forests, genomic identity-by-state methods) to build predictive models of these clinical phenotypes to identify at-risk individuals. (End of Abstract)