首页> 外文学位 >QTL variance component models.
【24h】

QTL variance component models.

机译:QTL方差分量模型。

获取原文
获取原文并翻译 | 示例

摘要

Variance components models are powerful and flexible tools for extracting genetic information and for mapping genes from quantitative trait data in pedigrees. We extend these models in two directions: multivariate outbred (human) analysis and inbred strain analysis. In the first direction, correlated traits exist, and it is uncertain which trait is primary as in the case of type 2 diabetes where, for example, obesity, impaired glucose tolerance and high cholesterol occur together. Sadly, only univariate or bivariate trait analysis is routinely carried out even when many traits have been measured because of a lack of appropriate software. Creating software is difficult for two mathematical reasons: one, estimating covariance matrices requires satisfying increasingly complicated constraints as the number of traits increases and two, estimating parameters requires maximizing a high-dimensional likelihood surface. By combining factor analysis with the traditional quantitative trait locus (QTL) mapping model and placing it in an efficient quasi-Newton algorithm, our software maps QTL simultaneously for any number of traits while providing information on pleiotropic major genes or polygenes and common environmental determinants with reasonable computation times. In the second direction, together with human studies, animal models of disease and QTL mapping studies are proceeding at a rapid pace. In the arena of inbred strains, cross-specific fixed effects models have been successful in mapping many genes using two strains, although cloning efforts lag far behind. To close the gap, many researchers are exploring multiple strain methods and observing the importance of background genetic variation. We have developed two new QTL mapping models that explicitly model the background genetic variation as a random effect which lead to general covariance expressions that are cross-independent and accomodate any number of strains. This generalization of the traditional QTL mapping model for outbred populations must take into account the peculiar genetic architecture of inbred strains: completely homozygous and genetically identical.
机译:方差组件模型是功能强大且灵活的工具,可用于提取遗传信息并从谱系中的定量性状数据中绘制基因。我们在两个方向上扩展了这些模型:多变量近交(人类)分析和近交菌株分析。在第一方向上,存在相关的性状,并且不确定哪个性状是主要的,例如在肥胖,葡萄糖耐量降低和高胆固醇同时发生的2型糖尿病的情况下。可悲的是,即使由于缺乏适当的软件而已测量了许多特征,也只能常规地执行单变量或双变量特征分析。由于两个数学原因,创建软件很困难:其一,随着特征数量的增加,估计协方差矩阵需要满足日益复杂的约束;其二,估计参数需要最大化高维似然面。通过将因子分析与传统的数量性状基因座(QTL)映射模型相结合,并将其置于高效的拟牛顿算法中,我们的软件可以同时对任意数量的性状进行QTL映射,同时提供有关多效性主要基因或多基因以及常见环境决定因素的信息。合理的计算时间。在第二个方向上,与人类研究一起,疾病的动物模型和QTL作图研究也在迅速进行。在近交菌株领域,尽管克隆工作远远落后,但交叉特异性固定效应模型已成功使用两种菌株绘制了许多基因。为了缩小差距,许多研究人员正在探索多种菌株方法,并观察背景遗传变异的重要性。我们已经开发了两个新的QTL映射模型,这些模型显式地将背景遗传变异建模为随机效应,从而导致交叉无关的通用协方差表达并适应任何数量的菌株。传统QTL映射模型对近交群体的这种概括必须考虑近交菌株的独特遗传结构:完全纯合和遗传相同。

著录项

  • 作者

    Bauman, Lara Elizabeth.;

  • 作者单位

    University of California, Los Angeles.;

  • 授予单位 University of California, Los Angeles.;
  • 学科 Biostatistics.;Bioinformatics.;Genetics.
  • 学位 Ph.D.
  • 年度 2007
  • 页码 117 p.
  • 总页数 117
  • 原文格式 PDF
  • 正文语种 eng
  • 中图分类
  • 关键词

相似文献

  • 外文文献
  • 中文文献
  • 专利
获取原文

客服邮箱:kefu@zhangqiaokeyan.com

京公网安备:11010802029741号 ICP备案号:京ICP备15016152号-6 六维联合信息科技 (北京) 有限公司©版权所有
  • 客服微信

  • 服务号