Bis J C, Glazer N L, Psaty B M
Department of Medicine, University of Washington Center for Health Studies, Group Health, Seattle, Washington, USA.
J Thromb Haemost. 2009 Jul;7 Suppl 1:308-11. doi: 10.1111/j.1538-7836.2009.03392.x.
Relying on known biology, candidate-gene studies have been only modestly successful in identifying genetic variants associated with cardiovascular risk factors. Genome-wide association (GWA) studies, in contrast, allow broad scans across millions of loci in search of unsuspected genetic associations with phenotypes. The large numbers of statistical tests in GWA studies and the large sample sizes required to detect modest-sized associations have served as a powerful incentive for the development of large collaborative efforts such as the Cohorts for Heart and Aging Research in Genomic Epidemiology (CHARGE) Consortium. This article uses published data on three phenotypes, fibrinogen, uric acid, and electrocardiographic QT interval duration, from the CHARGE Consortium to describe several methodologic issues in the design, conduct, and interpretation of GWA studies, including the use of imputation and the need for additional genotyping. Even with large studies, novel genetic loci explain only a small proportion of the variance of cardiovascular phenotypes.
基于已知生物学知识，候选基因研究在识别与心血管危险因素相关的基因变异方面仅取得了一定程度的成功。相比之下，全基因组关联（GWA）研究允许对数百万个位点进行广泛扫描，以寻找与表型的意外基因关联。GWA研究中的大量统计检验以及检测中等效应大小关联所需的大样本量，有力地推动了诸如基因组流行病学心脏与衰老研究队列（CHARGE）联盟等大型合作项目的开展。本文利用CHARGE联盟已发表的关于纤维蛋白原、尿酸和心电图QT间期持续时间这三种表型的数据，描述了GWA研究在设计、实施和解释中的几个方法学问题，包括插补的使用以及额外基因分型的必要性。即使是大型研究，新发现的基因位点也仅解释了心血管表型变异的一小部分。
