Genome-wide Haplotype Association Analysis in Mental Disorders

描述(申请人提供)：精神分裂症和抑郁症等精神疾病是一种复杂的疾病，其易感性、发展和治疗反应受到复杂的遗传和环境因素的调节。了解这些疾病的遗传学可以为这些疾病的诊断学、发病机制和治疗学的发展提供重要的见解。随着综合基因组信息和高性价比基因分型技术的发展，全基因组关联研究已成为识别复杂疾病适度遗传决定因素的一种很有前途的新工具。然而，针对精神障碍的GWA尚未带来确凿的结果。检测小效应基因的能力不足和无法纳入复杂的相互作用是缺乏可复制发现的两个主要属性。为了确保全球气候变化网络的进一步成功，需要先进的分析工具和战略来解决这些问题。这项提案打算根据这一需要制定方法。我们的长期目标是提高复杂多标记物分析的有效性，并最终促进研究设计和标记物选择。对多标记多态进行建模可以提供最大数量的基因组信息，而复杂的统计建模允许对潜在的遗传和环境因素进行仔细和集体的考虑。在这个方案中，我们关注基于模型的单倍型分析，并提出了一个两阶段框架来检测和理解GWAS中的关联信号。第一阶段旨在有效地筛选出具有全球单倍型-性状关联的地区。第二阶段侧重于更系统地研究单倍型效应的具体模式。受研究人员合作工作中出现的问题的激励，我们方法开发的主要考虑因素包括：(A)单倍型信息的有效利用，(B)基于回归的框架的制定，(C)检测主要和相互作用影响的能力，(D)从最初的全球筛选到后续具体评估的系统推断过程，以及(E)在用户友好的软件中建立坚实的理论基础和强有力的实施。我们将通过以下三个具体目标来实现我们的目标：(1)开发基于回归的单倍型相似性方法，用于检测显示遗传主效应和/或交互效应的区域；(2)开发惩罚似然回归方法，用于表征所识别区域内显著的主效应和/或交互效应的单倍型；以及(3)将AIMS(1)和(2)的方法应用于精神障碍的协作性GWAS，用于方法评估和疾病基因检测，并开发和分发供公众使用的计算机程序。 公共卫生相关性：拟议工作的完成将为研究复杂疾病的遗传病因学的新进程提供有效的统计工具，从最初的基因组筛查到随后的解释性检查。这些工具可以促进科学家对复杂疾病的理解，并最终导致更好地设计预防、检测和治疗战略，以改善人类健康。

DESCRIPTION (provided by applicant): Mental diseases such as schizophrenia and depression are complex diseases for which susceptibility, development and treatment response are mediated by intricate genetic and environmental factors. Understanding the genetics of these diseases can illuminate significant insights into the development of diagnostics, pathogenesis and therapeutics of these diseases. With recent advancements in comprehensive genomic information and cost-effective genotyping technologies, genome-wide association studies (GWAS) have become a promising new tool for identifying modest genetic determinants of complex disorders. However, GWAS for psychiatric disorders have yet to bring definitive findings. Insufficient power to detect small-effect genes and inability to incorporate complex interactions are the two major attributes for lack of replicable findings. To ensure further success of GWAS, advanced analytical tools and strategies are needed to resolve these issues. This proposal intends to develop methodology in response to this need. Our long-term goal is to advance the efficacy of complex multimarker analysis and eventually to facilitate study design and marker selection. Modeling multimarker polymorphisms provides maximal amount of genomic information, and complex statistical modeling allows careful and collective consideration of potential genetic and environmental factors. In this proposal we focus on model-based haplotype analysis, and propose a two-stage framework to detect and to comprehend the association signals in GWAS. The first stage aims to effectively screen out regions with global haplotype-trait association. The second stage focuses on a more systematic examination of the specific patterns of haplotype effects. Motivated by issues arising in the collaborative works by the investigators, the central considerations of our methodology development include: (a) the efficient usage of haplotype information, (b) the formulation of regression-based framework, (c) the capacity to detect main and interaction effects, (d) a systematic inference progression from initial global screening to follow-up specific evaluation, and (e) the establishment of a solid theoretical foundation and robust implementation in user-friendly software. We will achieve our objectives through the following three specific aims: (1) to develop regression-based haplotype-similarity methods for detecting regions that exhibit genetic main and/or interaction effects, (2) to develop a penalized-likelihood regression approach for characterizing haplotypes of significant main and/or interaction effects within the identified regions, and (3) to apply the methods from aims (1) and (2) to the collaborative GWAS of mental disorders for method evaluation and disease gene detection, and to develop and distribute computer programs for public use. PUBLIC HEALTH RELEVANCE: Completion of the proposed work will provide effective statistical tools for a new process of studying the genetic etiology of complex diseases, from initial genome screening to subsequent explanatory examination. These tools can facilitate scientists' understanding of complex diseases and eventually lead to better design of prevention, detection and treatment strategies to improve human health.