Statistical Methods in Genetic Studies of Substance Use

描述（由申请人提供）：从NIH指南PA-02-112“物质使用障碍的遗传流行病学”的出版中可以看出，物质使用障碍（SUDs）构成了严重的公共卫生和社会问题。基因-基因和基因-环境相互作用是了解sud遗传流行病学的潜在重要因素。在此过程中，先进的数据分析和建模技术不可或缺。分析的进步通常被认为是复杂疾病遗传解剖的最大挑战之一。为了满足这一需求，NIH指南PA-02-112鼓励发展统计工具和分析方法，以提高我们对包括sud在内的复杂表型的理解。

DESCRIPTION (provided by applicant): As evident from the publication of NIH Guide PA-02-112 "Genetic Epidemiology of Substance Use Disorders," substance use disorders (SUDs) pose serious public health and societal problems. Gene-gene and gene-environment interactions are potentially important factors in understanding the genetic epidemiology of SUDs. Advanced data analysis and modeling techniques become indispensable in this endeavor. Analytic advancement is generally regarded as one of the greatest challenges in genetic dissection of complex disease. To address this need, NIH Guide PA-02-112 encourages development of statistical tools and analytic methods to enhance our understanding of complex phenotypes including SUDs. This application response to this specific programmatic need by developing statistical methods that is useful for unraveling the genetic basis of complex disorders. Specifically, our primary aim is to develop, evaluate, and apply new statistical models (e.g., latent variable models and tree-based models), methods, and software to conduct genetic analyses of complex traits. We are particularly interested in ordinal traits because methods and software virtually do not exist. Once the methodologies are established, companion software will be developed for all of these models and made available to the public on Dr. Zhang's website. While the methodologies are being developed, as Dr. Zhang's group has demonstrated in the past, we will apply them to real data to address important public health problems, for example, those pertinent to the objectives of NIH Guide PA-02-112. We include three databases, which will allow us to study a variety of issues. For example, we will examine the potential sex difference in familial transmission of drug use and alcoholism, and identify candidate genes, gene-gene and gene-environment interactions for nicotine dependence. In our analyses, we will consider multiple phenotypes including alcohol, drug, and tobacco use as well as comorbid psychiatric conditions such as anxiety. Although the focus of application is on genetic epidemiology of SUBs, our methodologies will be useful for understanding complex ordinal traits in general.