基于高阶SNP互作挖掘与分析的复杂疾病全基因组关联研究

Complex disease refers to the disease caused by the common effect of multiple genes, which is a serious threat to human health, and brings a heavy burden to the family and society. Whole-genome association study based on the interactions of SNPs has opened a new chapter in the study of complex diseases, and has aroused the great enthusiasm of researchers. Although a number of methods have emerged to explore the interactions of SNPs in high-throughput genomic data, the severe computational challenge results in a lack of specialized higher-order interaction detection methods. There are many problems in the existing methods, such as multiple comparison, computational efficiency, provenance analysis, results integration and so on. This study focuses on the design of a series of novel methods for fast and accurate mining and analyzing of the high-order interactions in massive SNPs. The main contents of this research can be summarized as "four items", i.e. fast and precise correction of multiple comparison, complete and nil coupling parallel mining for high-order interactions, automation of tracing the cause of significant results, networking integration of the fragmented knowledge. The innovation points of this study are as follows: lightweight and precise permutation test based on flexible lower bound, parallel mining framework of high-order interactions based on maximal closed decomposition, tracing analysis of the significant interactions based on master-slave enumeration tree, disease pathway discovery based on locally heaviest subgraph. These research results can provide a new insight into the complex disease pathogenesis and accurate personalized medicine. Therefore, this research is of high theoretical values and extensive applications.

复杂疾病是多个基因共同作用导致的疾病，严重威胁着人类健康，给家庭和社会带来沉重的负担。基于SNP交互作用的全基因组关联分析开启了复杂疾病研究的新篇章，激发了众多研究者极大的热情。虽然涌现出一批探索高通量基因组学数据SNP交互作用的方法，但严峻的计算挑战，致使专门的高阶交互作用检测方法匮乏。现有方法存在多重比较、计算效率、溯源分析、结果整合等方面的问题。本课题侧重于设计面向海量SNP的快速精准高阶互作挖掘与分析新方法。主要研究内容概括为“四化”：多重比较精准校正的快速化、互作挖掘的完备零耦合并行化、显著互作溯源分析的自动化、碎片知识系统整合的网络化。研究中，拟在基于柔性下界的轻量级精确置换检验技术、基于极大闭分解的并行高阶互作挖掘框架、基于主-从枚举树的显著互作溯源分析方法、基于局部最重子图的疾病通路网发现算法上有所创新。研究成果对揭示复杂疾病发病机理，精准个性化医疗等，具有非常重要的意义。

复杂疾病是多个基因共同作用导致的疾病，严重威胁着人类健康，给家庭和社会带来沉重的负担。基于SNP交互作用的全基因组关联分析开启了复杂疾病研究的新篇章，激发了众多研究者极大的热情。虽然涌现出一批探索高通量基因组学数据SNP交互作用的方法，但严峻的计算挑战，致使专门的高阶交互作用检测方法匮乏。现有方法存在多重比较、计算效率、溯源分析、结果整合等方面的问题。本课题侧重于设计面向海量SNP的快速精准高阶互作挖掘与分析新方法。主要研究内容概括为“四化”：多重比较精准校正的快速化、互作挖掘的完备零耦合并行化、显著互作溯源分析的自动化、碎片知识系统整合的网络化。研究中，拟在基于柔性下界的轻量级精确置换检验技术、基于极大闭分解的并行高阶互作挖掘框架、基于主-从枚举树的显著互作溯源分析方法、基于局部最重子图的疾病通路网发现算法上有所创新。项目取得的研究成果对揭示复杂疾病发病机理，精准个性化医疗等，具有非常重要的意义。

内容获取失败，请点击重试