多位点基因关联分析

发布时间：2019-01-05 02:04

【摘要】：为疾病需找分子依据一直以来都是人类的梦想,随着人类基因组计划的完成与基因测序技术的高速发展,这一梦想即将成为现实。单核苷酸多态性(Single Nucleotide Polymorphisms, SNP)是当前人类基因关联分析研究中的常用标记物,目前,基于单SNP位点的关联分析研究已投入实际使用。但是,由于人类基因组规模庞大,而且复杂疾病往往是由多个基因共同确定的,因而这类单位点研究方法已远远不能满足需要。寻求一种高效而可靠的多位点基因关联分析方法已成为人们争相研究的热点。目前,科学家们已经从多个角度出发,提出了多种关联分析检验方法,但由于基因变量的分布规律以及SNP位点间连锁不平衡性的存在,这些方法往往存在各种缺陷,仍未找到一种可大规模应用的方法。本文回顾了近年来一些流行的关联分析方法,尤其是近年来发展迅速的相似度方法,并从三个不同角度出发,提出了三种新方法,分别是基于Logistic回归的方法,基于多元方差分析的方法以及基于潜在正态变量模型的方法,最后通过模拟实验将这些方法与传统方法进行了比较。在模拟实验中,我们分别建立了独立基因与带有连锁不平衡的基因模型,并调整了易感位点数量,背景连锁不平衡水平等多种参数以模拟不同的基因环境,模拟结果表明,在各种条件下,相比传统方法,我们的检验方法有着较强的功效。最后,在未来的研究中,我们的方法可以由二项性状推广至复杂性状,有较强的发展前景。
[Abstract]:It has always been a human dream to find molecular basis for disease. With the completion of the human genome project and the rapid development of gene sequencing technology, this dream will become a reality. Single nucleotide polymorphic (Single Nucleotide Polymorphisms, SNP) (SNP) is a common marker in human gene association analysis. At present, the association analysis based on single SNP locus has been put into practical use. However, due to the large scale of human genome and the fact that complex diseases are often determined by multiple genes, this kind of unit point research method is far from satisfying the needs. Searching for an efficient and reliable method for multilocus gene association analysis has become a hot topic. At present, scientists have put forward a variety of association analysis methods from many angles. However, due to the distribution of genetic variables and the existence of linkage disequilibrium between SNP loci, these methods often have various defects. A method that can be applied on a large scale has not yet been found. This paper reviews some popular association analysis methods in recent years, especially the rapidly developing similarity analysis methods in recent years. Three new methods are proposed from three different angles, which are based on Logistic regression. The methods based on multivariate analysis of variance and potential normal variable model are compared with the traditional methods through simulation experiments. In the simulation experiment, we established independent gene model and gene model with linkage disequilibrium, and adjusted the number of susceptible sites, background linkage disequilibrium level and other parameters to simulate different gene environment. Under various conditions, compared with the traditional method, our test method has a strong effect. Finally, in the future research, our method can be extended from binomial traits to complex traits.
【学位授予单位】：华东师范大学
【学位级别】：硕士
【学位授予年份】：2012
【分类号】：R346

【相似文献】