Integrated Variation Detection Annotation and Analysis

描述（由申请人提供）：正在以前所未有的速度产生关于多种物种基因组的高通量测序（HTS）数据。然而，处理这些数据的计算和统计方法的发展滞后，在产生的大量数据和可以收集的生物知识之间产生了差距。在这里，我们建议开发一个集成的系统，用于HTS数据的遗传变异检测，注释和分析，从而减少社区面临的关键差距。在目标1中，我们将开发一种基于隐马尔可夫模型（HMM）的计算算法，该算法结合了多个信息源，包括序列深度，等位基因剂量，群体等位基因频率和配对末端读数距离，用于可靠而有效地检测拷贝数变异（CNV）。鉴于大量SNP、插入缺失和CNV，研究人员面临着识别功能重要变体子集的挑战。在目标2中，我们将开发一个全面的功能注释管道来注释编码和非编码变体的功能重要性，利用来自许多大规模基因组学项目的数据库信息，并为每个变体生成“功能向量”。这些功能载体可以帮助生物学家解释测序结果，并帮助统计遗传学家使用测序数据开发知情关联测试。需要适当的统计方法来分析群体水平的测序数据，以识别可能导致疾病易感性或表型变异的基因组变异。在目标3中，我们将开发一种分层建模策略，该策略利用每个变体的功能向量信息，对基因、基因组区域或生物途径（如本体类别和基因调控/代谢途径）进行关联测试。最后，在目标4中，我们将通过模拟和真实的数据分析来测试每种方法的性能，并开发、分发和支持实现所提出方法的免费软件包。我们相信，有据可查和支持的软件实现将使其他研究人员能够从该项目产生的方法和科学进步中获得最大的信息。这些目标的成功完成将使研究人员能够充分研究已经或将要产生的大量测序数据，从而有助于我们了解遗传变异如何影响表型变异。

DESCRIPTION (provided by applicant): High-throughput sequencing (HTS) data on the genomes of a diverse number of species are being produced at an unprecedented rate. However, the development of computational and statistical approaches for handling these data lags behind, creating a gap between the massive data being generated and the biological knowledge that could be gleaned. Here we propose to develop an integrated system for genetic variation detection, annotation and analysis for HTS data, therefore reducing the critical gap faced by the community. In Aim 1, we will develop a hidden Markov model (HMM) based computational algorithm that incorporates multiple sources of information, including sequence depth, allelic dosage, population allele frequency and paired-end reads distance, for reliable yet efficient detection of copy number variations (CNVs). Given a large list of SNPs, indels and CNVs, researchers are faced with the challenge of identifying a subset of functionally important variants. In Aim 2, we will develop a comprehensive functional annotation pipeline to annotate functional importance of coding and non-coding variants, utilizing database information from many large-scale genomics projects, and generate a "functional vector" for each variant. These functional vectors can help biologists interpret sequencing results and help statistical geneticists develop informed association tests using sequencing data. Appropriate statistical methods are needed to analyze population-level sequencing data, in order to identify genomic variants that may contribute to disease susceptibility or phenotypic variability. In Aim 3, we will develop a hierarchical modeling strategy, which utilizes functional vector information for each variant, to perform association tests on genes, genomic regions, or biological pathways, such as ontology categories and gene regulatory/metabolic pathways. Finally, in Aim 4, we will test the properties of each approach via simulation and real data analysis, and develop, distribute and support freely available software packages implementing the proposed methods. We believe that well-documented and supported software implementations will allow other researchers to yield the maximum information from the methodological and scientific advances that result from this project. Successful completion of the aims will enable researchers to fully investigate the massive amounts of sequencing data that have been or will be generated, thus contributing to our understanding on how genetic variants influence phenotype variability.