Identification of Rare Alleles for Genetic Association and Risk

描述(申请人提供)：全基因组关联研究最近被应用于检测导致人类疾病易感性的基因变异，如糖尿病、心脏病和癌症。通过设计，这些关联可能是间接获得的，因为可能没有观察到影响疾病的基因类型。相反，人们希望遗传风险因素可能与观察到的基因数据充分相关，从而建立关联。捕捉常见变异的标准标记集在检测风险变异也常见的关联方面非常强大，但影响疾病表型的罕见遗传变异可能仍未被检测到。然而，特定的基因类型组合可以有效地将基因类型标记在一个风险易感的遗传位点上，从而促进关联的检测。虽然邻近基因座上的基因类型之间的相关性对于检测关联是必不可少的，但在试图确定关联的最终来源，即致病遗传基因座时，它们是繁重的。在这项应用中，我们概述了一种基于单倍型的统计方法，用于检测和剖析二元表型和罕见遗传变异之间的关联。此外，我们利用我们的模型来识别风险等位基因的个体携带者。这既有助于描述基因对复杂表型的影响，也为制定初步的风险模型提供了工具。我们将把这些方法应用于肺癌全基因组关联研究的数据。我们的方法在计算上是容易处理的，适用于现有和即将到来的大型数据集，将被纳入广泛使用和免费提供的FAST PHASE软件包中。 公共卫生相关性：确定复杂人类疾病的遗传原因需要有效利用全基因组变异调查的可用数据。单倍型变异提供了信息，以检测罕见遗传因素对疾病的影响，并可以特别有助于预测哪些人患疾病的风险最高。

DESCRIPTION (provided by applicant): Genome-wide association studies have recently been applied to detect genetic variants that contribute to the predisposition of human disease, such as diabetes, heart disease, and cancer. By design, these associations may be obtained indirectly, as the disease-influencing genotypes may not be observed. Rather, the hope is that genetic risk factors may be sufficiently correlated with observed genotype data, with which an association may be established. Standard marker sets, which capture common variation, are powerful to detect associations where the risk variants are also common, but rare genetic variants that influence disease phenotypes may remain undetected. However, specific combinations of genotypes can effectively "tag" the genotypes at a risk- predisposing genetic locus, facilitating the detection of association. While the correlations among genotypes at nearby loci are essential for detection of an association, they are burdensome when trying to identify the ultimate sources of association, i.e. causative genetic loci. In this application, we outline a haplotype-based statistical approach to detecting and dissecting association between a binary phenotype and rare genetic variants. In addition, we utilize our model to identify individual carriers of the risk alleles. This serves both to aid in the characterization of the genetic influence on a complex phenotype, as well as to provide a tool for formulating preliminary risk models. We will apply these methods to data from a genome-wide association study of lung cancer. Our methods, which are computationally tractable for the application to large existing and forthcoming data sets, will be incorporated into the widely used and freely available fastPHASE software package. PUBLIC HEALTH RELEVANCE: Identification of the genetic causes of complex human disease requires efficient use of available data from genome-wide surveys of variation. Haplotype variation provides information to detect influence on disease from rare genetic factors, and can be particularly helpful in predicting which individuals are at highest risk for developing disease.