Integrative Approaches to Mapping Susceptibility Genes of Complex Neuropsychiatric Disorders

项目摘要 识别神经发育和精神疾病的易感基因和变体不仅 有助于我们理解这些疾病，但也指出了潜在的治疗目标。全基因 关联研究（GWAS）通常用于研究复杂疾病，包括神经精神疾病 疾病然而，GWAS专注于常见的变体，并且在研究早期- 发病疾病，包括许多发育障碍，其风险等位基因通常保持在非常低的水平 人口的频率。此外，GWAS的结果通常不能直接转化为知识 风险基因和疾病机制。 该项目的目标是开发综合的统计方法来分析遗传数据， 神经精神疾病，以绘制其易感基因，并获得疾病遗传学的见解。(1)我们 提出分析患者家族外显子组和基因组测序数据的方法。不同于现有 遗传学研究的方法通常集中在每次的数据类型上，我们的方法将整合广泛的 基因水平的遗传变异谱，包括非同义和调控性非编码 突变，包括从头突变和从父母遗传的突变。这导致更高的检测风险的能力 基因. (2)拷贝数变异（CNVs）在神经发育障碍中起重要作用。但 CNVs通常与多个基因重叠，并且很难在疾病相关的CNVs中识别风险基因。一个新 提出了一种从CNVs中提取基因级信息的算法。这使我们能够将联合收割机CNV数据和 测序中的核苷酸变异数据，以更好地检测疾病基因。(3)非编码的重要性 复杂疾病的变异现已牢固建立，表达QTL（eQTL）是一个有前途的 战略映射非编码变异有功能的基因表达水平的影响。我们提出了一个 eQTL和GWAS数据联合分析的新统计方法。该方法的独特之处在于它使用 一个基因的所有eQTL信息，以测试其在疾病中的作用，包括顺式和反式eQTL，跨 整个效应量范围。(4)我们努力的一个关键组成部分是将我们开发的方法整合到 用户友好的软件，可以使广泛的精神病遗传社区受益。

Project Summary Identifying the susceptibility genes and variants of neurodevelopmental and psychiatric diseases will not only contribute to our understanding of these diseases, but also point to potential therapeutic targets. Genome-wide association studies (GWAS) are commonly used to study complex diseases, including neuro-psychiatric diseases. Nevertheless, GWAS focus on common variants, and have not been successful in studying early- onset diseases, including many developmental disorders, whose risk alleles are generally kept at very low frequencies in population. Additionally, the results of GWAS often cannot be directly translated into knowledge of risk genes and disease mechanisms. The goal of this project is to develop comprehensive statistical methods for analyzing genetic data of neuropsychiatric diseases to map their susceptibility genes and gain insights of the disease genetics. (1) We propose methods to analyze exome and genome sequencing data from patient families. Unlike existing methods for genetic studies which often focus on type of data per time, our methods will integrate a broad spectrum of genetic variations at the level of genes, including non-synonymous and regulatory non-coding mutations, both de novo and inherited from parents in origin. This leads to a higher power of detecting risk genes. (2) Copy number variants (CNVs) make substantial contribution to neurodevelopmental disorders. But CNVs often overlap multiple genes and it is difficult to identify risk genes within disease-related CNVs. A new algorithm is proposed to extract gene-level information from CNVs. This allows us to combine CNV data and nucleotide variation data from sequencing, to better detect disease genes. (3) Importance of non-coding variants to complex disease has now been firmly established and expression QTL (eQTL) is a promising strategy to map non-coding variants that have functional effects on gene expression levels. We propose a novel statistical approach to joint analysis of eQTL and GWAS data. The method is unique in that it uses information of all eQTL of a gene to test its role in disease, including both cis- and trans-eQTL, across the entire range of effect sizes. (4) A key component of our effort is the integration of the methods we develop into user-friendly software that could benefit the broad psychiatric genetic community.