人类群体遗传空间结构异质性理论与定量分析方法研究

发布时间：2018-05-08 23:27

本文选题：人类群体遗传学 + 人类群体遗传空间结构　；参考：《山东大学》2005年博士论文

【摘要】：人类群体遗传结构是指人群中所有个体的全部基因座的基因频率。目前,研究人类群体遗传结构的空间结构模式及其变化规律的定量分析方法,在分析群体遗传结构的空间特征、空间估计、群体空间关系和基因流动轨迹、遗传界线的识别、遗传结构空间成分的定量剖分等方面均存在局限性。为此,本研究遵循人类群体遗传学原理,将地质统计学、数学生态学、混沌分形理论、图论、分子遗传学等学科理论方法相结合,研究了人类群体遗传空间结构异质性(即复杂性和变异性)的定量分析方法,旨在完善人类群体遗传空间结构定量分析方法体系。全文由以下六章内容组成: 第一章基因选择与资料收集:根据不同类型基因座的突变、进化机制和收集的群体遗传学资料,选择了趋化因子CCR2、趋化因子受体SDF-1、ABO、HLA—A,B、TPOX、FGA、CSF1PO、D7S820、THO1和VWA基因座的多态性资料作为本研究的基本数据,构建了各基因座的空间数据库。第二章人类群体遗传结构的理论模型和常用测度指标:简介了人类群体遗传结构的有关理论模型(岛屿模型、步石模型和距离隔离模型)和常用测度指标(基因频率及其方差协方差、杂合度和基因多样性、遗传分化系数和固定指数、遗传距离),以便于理解群体遗传结构的有关概念。第三章人类群体遗传空间结构分析的经典多元统计模型及其改进:从分析基因频率矩阵结构特点入手,探讨了应用经典多元统计模型分析群体遗传空间结构时所存在的问题及其改进方法:①针对基因频率矩阵的“闭合”特点,提出了对数比非线性多元分析方法;②通过对提取基因频率矩阵主成分方法的比较,证明均值化和中心化协方差阵是提取基因频率矩阵主成分的有效方法;③利用主成分特征根的权重性和特征向量对群体遗传结构变异性作用方向的信息,定义了主成分综合遗传效应测度(SPC);④将图论和多元统计方法有机结合,构建了既可反映群体遗传空间结构特性,又可揭示群体问遗传学联系的“图论多元排序分析”模型;⑤通过分析群体遗传空间结构对应分析中“蹄型效应”产生的原因,提出了其校正方法;⑥构建了人类群体遗传空间结构的PPG Biplot模型,其几何性质和特征参数反映了丰富的群体遗传学含义。
[Abstract]:The genetic structure of human population refers to the gene frequency of all loci of all individuals in the population. At present, the quantitative analysis method to study the spatial structure pattern and its changing law of human population genetic structure is used to analyze the spatial characteristics of population genetic structure, the spatial estimation, the spatial relationship of population, the locus of gene flow, and the identification of genetic boundary. There are some limitations in the quantification of spatial components of genetic structures. Therefore, following the principle of human population genetics, this study combines the theories of geostatistics, mathematical ecology, chaos fractal theory, graph theory, molecular genetics and so on. The quantitative analysis method of genetic spatial structure heterogeneity (i.e. complexity and variability) of human population is studied in order to perfect the quantitative analysis system of human population genetic spatial structure. The full text consists of the following six chapters: Chapter I Gene selection and data Collection: based on mutations in different types of loci, evolutionary mechanisms and collected population genetic data, The polymorphism data of chemokine CCR2, SDF-1ABOOG HLA-APOXFGAA, CSF1POD7S820THO1 and VWA loci were selected as the basic data of this study, and the spatial database of each locus was constructed. The second chapter is the theoretical model of the genetic structure of human population and the commonly used measure index: the related theoretical models of the genetic structure of human population (island model, island model) are briefly introduced. Stepstone model and distance isolation model) and commonly used measures (gene frequency and variance covariance, heterozygosity and gene diversity, genetic differentiation coefficient and fixed index, genetic distance), in order to understand the population genetic structure related concepts. In chapter 3, the classical multivariate statistical model of genetic spatial structure analysis of human population and its improvement are introduced. This paper discusses the problems existing in the analysis of population genetic spatial structure using classical multivariate statistical model and its improved method: 1. Aiming at the "closed" characteristics of gene frequency matrix, a logarithmic nonlinear multivariate analysis method is proposed. 2Compared with the method of extracting the principal component of gene frequency matrix, it is proved that mean value and centralization covariance matrix are effective methods to extract the principal component of gene frequency matrix; (3) based on the information of the weight of the principal component characteristic root and the effect of the characteristic vector on the population genetic structure variability, the paper defines the principal component comprehensive genetic effect measure and combines the graph theory with the multivariate statistical method. The "graph theory multivariate sequencing analysis" model, which not only reflects the characteristics of population genetic spatial structure, but also reveals the causes of "hoof effect" in population genetic spatial structure correspondence analysis, is constructed. The PPG Biplot model of the genetic spatial structure of human population is constructed by its correction method. The geometric properties and characteristic parameters of the model reflect the rich meaning of population genetics.
【学位授予单位】：山东大学
【学位级别】：博士
【学位授予年份】：2005
【分类号】：Q987

【引证文献】