Multi-point and multi-locus analysis of genomic association data

描述（由申请人提供）： 全基因组关联研究（GWAS）为研究遗传变异对复杂疾病风险的影响提供了一种新的强有力的方法。随着基因分型技术的进步，全基因组关联研究正在成为现实。来自GWAS的数据预计将以更快的速度增长。尽管在开发用于映射复杂疾病/性状的有效算法方面做出了巨大努力，但基于单位点的方法仍然是GWAS的主要方法。然而，众所周知，通常多种遗传因素、环境因素以及它们之间的相互作用在复杂疾病的病因学中起着重要作用。同时从数十万个单核苷酸多态性（SNP）的多个变量和它们的相互作用建模的新的和实用的方法是非常需要的。在这个项目中，我们建议开发有效的算法和实用的统计工具，以解决两个重要的问题，在全基因组关联研究的背景下：多点分析和多位点分析。对于多点分析，我们的动态隐链马尔可夫模型（DHCMM）可以联合建模的历史重组和穆塔- tions，单倍型结构和频率，和协会，这是预计将比现有的方法更有效。对于多位点分析，我们建议使用先进的机器学习方法来联合筛选预测疾病的SNP。我们的集成软件系统MAVEN将利用尖端技术促进全球水资源评估数据的管理、分析、可视化和结果共享。GWAS的真正价值取决于有效的计算模型和工具的开发。我们预计，该研究项目将大大加快对复杂疾病遗传结构的理解。

DESCRIPTION (provided by applicant): Genome-wide association studies (GWAS) provide a new and powerful approach to investigate the effect of inherited genetic variation on risks of complex diseases. With recent advances in genotyping technology, genome-wide association studies are now becoming a reality. Data from GWAS are expected in an accelerated rate. Despite tremendous efforts in developing efficient algorithms for mapping complex diseases/traits, single-locus based approaches are still the primary method for GWAS. However, it is known that usually multiple genetic factors, environmental factors as well as their interactions play an important role in the etiology of complex diseases. Novel and practical approaches to simultaneously model multiple variables and their interactions from hundreds of thousands single nucleotide polymorphisms (SNPs) are greatly needed. In this project, we propose to develop efficient algorithms and practical statistical tools to address two important problems in the context of genome- wide association studies: multi-point analysis and multi-locus analysis. For multi-point analysis, our Dynamic Hidden Chain Markov Model (DHCMM) can jointly model historical recombination and muta- tions, haplotype structures and frequencies, and associations, which is expected to be more effective than existing approaches. For multi-locus analysis, we propose to use an advanced machine learning approach to jointly screen SNPs that are predictive of diseases. Our integrated software system MAVEN will facilitate management, analysis, visualization and results sharing of GWA data using cut- ting edge technologies. The true value of GWAS is pending the development of effective computational models and tools. We anticipate that this research project will greatly accelerate the understanding of the genetic architecture of complex diseases.