基于通路的全基因组关联分析策略应用于西门塔尔牛的初步研究

发布时间：2018-03-06 15:38

本文选题：算法　切入点：西门塔尔牛　出处：《中国农业科学院》2015年硕士论文　论文类型：学位论文

【摘要】：随着人类全基因组计划的完成及高通量基因分型技术的快速发展,全基因组关联分析(GWAS)已经成为挖掘人类疾病易感基因及与家畜动物重要经济性状相关QTL的有力工具。大量以单核苷酸多态性片段(SNP)为主遗传标记已经被发现与人类疾病和家畜重要经济性状显著相关。然而,随着研究的深入,人们逐渐发现GWAS检测到的显著性SNP位点只能解释小部分与表型相关的变异,大量的剩余遗传信息还没有被挖掘。此外,大多数复杂的数量遗传性状往往由微效多基因决定的,单个SNP无法解释所有的表型变异。因此,需要提出新的统计策略来解决这种缺陷。在已有的众多策略中,基于通路的全基因组关联分析方法评估处于同一生物学通路中多个SNP的共同作用,而不仅仅单独的分析单个SNP的效应而广泛受到人们的关注。一些基于通路的关联分析算法也相继被提出,然而传统的基于通路的全基因组分析方法仅仅简单的使用显著的SNP构建基因统计量,忽视了小效应SNP的作用,这种策略有很大的局限性。为了解决传统的基于通路分析的算法缺陷,本研究提出了一种新的算法。这种算法包括两个步骤:1)使用主成分分析的思想构建每个基因的主成分矩阵,然后根据主成分与表型的关联程度对其进行排序。2)使用最大均值法(Maxmean)构建每个基因的统计量,然后使用柯尔莫哥罗夫-斯米尔诺夫(Kolmogorov-Smirnov)检验通路的显著性。基于来自内蒙古乌拉盖地区807头西门塔尔牛所包含的77,000,000 SNPs的数据,本研究使用新的算法对宰前活重、眼肌面积两个重要的经济性状进行了分析。结果表明,在对来自于KEGG数据库的263条通路的分析研究中,伽马氨基丁酸通路(p=0.000876)和非酒精性脂肪性肝病疾病(NAFLD)通路(p=0.000058)分别与宰前活重和眼肌面积两个性状显著相关。对两条显著性通路的生物信息学分析发现,伽马氨基丁酸通路已经被证实与动物的采食量及体重增加过程相关。本研究提出的基于通路的新算法首次剔除基因内部SNP之间的连锁作用,丰富了现有基于通路的全基因关联分析的算法。更重要的是,本项研究是首次对肉牛进行了基于通路的全基因组关联分析,共发现两条与肉牛重要经济性状相关的显著性通路。研究结果可能为后续实施分子育种提供重要的前提和基础。
[Abstract]:With the completion of the Human Genome Project and the rapid development of high-throughput genotyping technology, Genome-wide association analysis (GWAS) has become a powerful tool for mining human disease susceptibility genes and QTL associated with important economic traits in domestic animals. A large number of single nucleotide polymorphic fragments (SNPs) have been found to be associated with human genetic markers. Similar diseases are significantly associated with important economic traits in domestic animals. However, As the study progressed, it was gradually found that significant SNP loci detected by GWAS could only explain a small number of phenotypic variations, and that a large amount of residual genetic information had not yet been mined. Most complex quantitative genetic traits are often determined by microfunctional polygenes, and a single SNP cannot explain all phenotypic variations. Therefore, new statistical strategies are needed to address this defect. Pathway-based genome-wide association analysis was used to assess the interaction of multiple SNP in the same biological pathway. Not only the analysis of the effects of a single SNP has attracted widespread attention, but also a number of path-based association analysis algorithms have been proposed. However, the traditional genome-wide analysis method based on pathway simply uses significant SNP to construct gene statistics, neglecting the function of small effect SNP. This strategy has great limitations. In order to solve the shortcomings of traditional algorithms based on path analysis, This study proposes a new algorithm, which consists of two steps: 1) using the idea of principal component analysis (PCA) to construct a principal component matrix for each gene. And then sort it according to the association between the principal components and phenotypes. 2) using the maximum mean method Maxmean) to construct the statistics for each gene. Then using Colmogorov Smirnov Kolmogorov-Smirnov) to test the significance of the pathway. Based on the data of 77,000,000 SNPs from 807 Simmental cattle in Urage, Inner Mongolia, a new algorithm was used to measure the premortem weight. Two important economic traits of eye muscle area were analyzed. The results showed that, in the analysis of 263 pathways from KEGG database, The gamma-aminobutyric acid (GABA) pathway (0.000876) and the NAFLD (non-alcoholic fatty liver disease) pathway (p0.000058) were significantly correlated with pre-mortem weight and eye muscle area, respectively. The gamma-aminobutyric acid pathway has been shown to be related to animal intake and weight gain. It enriches the existing algorithms for the whole gene association analysis based on pathway. More importantly, this study is the first time that the whole genome association analysis based on pathway has been carried out in beef cattle. Two significant pathways related to important economic traits of beef cattle were found. The results may provide an important premise and basis for further implementation of molecular breeding.
【学位授予单位】：中国农业科学院
【学位级别】：硕士
【学位授予年份】：2015
【分类号】：S823

【相似文献】