Analysis of Genomic Data for Complex Traits

描述(申请人提供)：许多健康状况，包括物质使用和精神疾病，都是复杂的，取决于遗传和环境因素。在过去的几年里，全基因组关联研究(GWA)已经确定了单核苷酸多态，涉及数百个强健复制的共同特征的基因座。尽管已经取得了许多成功，但对于复杂疾病来说，识别基因、环境因素以及它们之间的相互作用仍然是困难的。这被称为遗传学家的噩梦。大多数已识别的变异具有较低的相关风险，几乎没有遗传性，人们越来越关注寻找复杂疾病的“缺失遗传性”。为此，重要的是开发新的统计方法。我们的初步进展表明，我们提出的方法已经在基因、环境和复杂性状之间的关联方面产生了重要的发现。我们用我们的新方法确定的几个遗传变异将由国家人类基因组研究所编目。这个项目将利用他在全球遗传研究数据收集和分析方面的多年经验，并在他成功开发基因研究的统计方法和软件的基础上再接再厉。这一应用程序的主要目标是继续努力并取得成功，开发、评估和应用新的统计模型、方法和软件来进行复杂疾病的GWA分析。我们的具体目标如下：(1)开发统计方法来对多维和多峰特征进行推理。将开发新的方法，通过纳入多种变种来发现隐藏的遗传力；同时考虑遗传和环境，并对多种和异质性状进行建模；(A2)通过纳入多种遗传变种、协变量和基因-协变量相互作用并纳入现有的生物信息，开发基于树和森林的关联分析方法；(A.3)开发和发布软件，通过PI的网站供公众使用。在开发方法和软件的同时，它们将被应用到各种实际研究中，作为我们方法和软件的动机和验证。在这方面，我们的次要目标是(B1)确定成瘾、精神疾病和精神障碍的共同发病率的基因和环境因素；(B2)确定早产的遗传变异和环境因素。简而言之，目标是 这个项目的意义重大，我们的方法的基础已经过测试，新的开发将是新颖和有用的。PI在这个项目上有几十年的经验，领导着一个拥有完善的基础设施和支持人员和学生的研究中心。 公共卫生相关性：尽管技术和方法的巨大进步导致了最近在识别复杂疾病的遗传变异方面取得的成功，但新的统计方法的发展在处理复杂表型的遗传研究中固有的困难方面至关重要。这个项目将对基因数据的分析产生重大影响，从而对公众健康产生重大影响，因为我们的方法和软件可以帮助研究人员了解常见和复杂疾病的遗传和环境因素，包括药物使用、癌症和早产。

DESCRIPTION (provided by applicant): Many health conditions, including substance use and mental illnesses, are complex and depend on both genetic and environmental factors. In the past several years genome wide association studies (GWA) have identified single-nucleotide polymorphisms implicating hundreds of robustly replicated loci for common traits. Despite numerous successes, it remains persistently difficult to identify genes, environmental factors, and interactions among them for complex diseases. This has been referred to as the geneticist's nightmare. Most of the identified variants have low associated risks and account for little heritability, and there is an increasing attention to find the "missing heritability" of complex diseases. To this end, it is important to develop novel statistical methods. Our Preliminary Progress demonstrates that our proposed methods have already produced significant findings on the association between genes, environments, and complex traits. Several genetic variants that we identified by our novel methods will be cataloged by National Human Genome Research Institute. This project will take advantage of the PI's many years of experience in the data collection and analysis of GWA studies and build on his success in the development of statistical methods and software for genetic studies. The primary aim of this application is to continue our effort and success in developing, evaluating, and applying new statistical models, methods, and software to conduct GWA analyses of complex diseases. Our specific aims are as follows: (A1) to develop statistical methods to perform inference for multidimensional and multi-modal traits. New methods will be developed to find the hidden heritability by incorporating multiple variants; simultaneously considering genetics and environment, and modeling multiple and heterogeneous traits; (A2) to develop tree- and forest-based methods for association analyses by incorporating multiple genetic variants, covariates, and gene-covariate interactions and incorporating existing biological information; (A.3) to develop and release software for public use through the PI's website. While the methods and software are developed, they will be applied to a variety of real studies that will serve as motivation and validation of our methods and software. In this regard, our secondary aims are to (B1) identify genes and environmental factors for addiction, mental illnesses, and the co-morbidity of psychiatric disorders; and (B2) identify genetic variants and environmental factors for preterm deliveries. In short, the objective of this project is significant, the foundation of our approach has been tested, and the new development will be novel and useful. The PI has decades of experience related to this project and leads a research center with well-established infrastructure and supporting personnel and students. PUBLIC HEALTH RELEVANCE: Despite great advances in technology and methodology that have led to recent successes in identifying genetic variants for complex diseases, developments of novel statistical methods are critically important in dealing with difficulties inherent in geneic studies of complex phenotypes. This project will have a significant impact on analysis of genetic data and hence on public health, because our methods and software can help investigators understand genetic and environmental factors of common and complex diseases including substance use, cancer, and preterm birth.