项目名称: 基于单体型的基因统计关联分析
项目编号: No.11271346
项目类型: 面上项目
立项/批准年度: 2013
项目学科: 数理科学和化学
项目作者: 杨亚宁
作者单位: 中国科学技术大学
项目金额: 60万元
中文摘要: 复杂疾病的基因关联分析关键在于寻找和发现多个基因在病例和正常人之间统计分布的不同。单体型作为同一条染色体上不同位点的等位基因组成的序列,是基因位点之间连锁不平衡(LD)或相依信息的有效载体,其分布在病例与对照组之间的差异反映了多个等位基因对于疾病的共同作用。基因型数据不提供单体型的相型信息,因而基因型数据的单体型分析是一种不完全数据的统计分析方法。本项目研究以单体型为基本单位的基因关联分析,包括利用群体中单体型的稀疏性研究单体型频率估计的压缩感知规则化算法;基于LD系数矩阵或复合LD系数矩阵主要特征在病例组和对照组之间的差异,构建交互作用分析方法;基于回溯型似然函数,通过组合划分方法降低参数空间维数,研究以单体型为基础的高效检验方法;利用下一代测序技术提供的大量低频率变异,研究贝叶斯方法以及单体型方法整合多个稀有变异的效应。本项目所研究的方法可应用于复杂疾病的全基因组多基因关联分析。
中文关键词: 关联分析;单体型;交互作用;复杂疾病;
英文摘要: Genetic association analysis aims at uncovering the differential pattern or distribution of genetic components between cases and controls for complex diseases. A haplotype is a sequence of alleles on the same chromosome, which is regarded as linkage disequilibrium(LD) information-rich carrier and an effective tool to study gene-gene interaction. Genotype data do not provide phase information and haplotypes are not observable, therefore genotype data are incomplete when haplotypes are the unit of the study. We propose to investigate haplotype-based association analysis methods, expecting that haplotypes as the unit of association analysis can provide insightful information about multi-gene interactions. We will study the regularization approach in compressive sensing theory to estimate haplotype frequencies, based on the fact that the existing haplotypes in population are rare; By contrasting the eigen-vectors and eigen-values of the LD matrices or composite LD matrices, we explore the interaction pattern in multi-locus analysis; By applying the combinatorial partitioning methods for retrospective likelihood function, we investigate the method of collapsing haplotypes to improve power of genotype-based association analysis; The next-generation sequencing method have provided much more common and rare genetic va
英文关键词: Association analysis;haplotype;interaction;complex disease;