论文部分内容阅读
通过将“克立格”技术应用于人类群体遗传学领域,构建了人类群体遗传空间结构的“克立格”模型,并论 述了其原理和计算方法。以HLA A基因座为例,应用“克立格”模型,定量分析了中国人群HLA A基因座的空间 遗传异质性;对HLA A基因频率的空间数据矩阵进行了主成分分析,进而定义了人类群体遗传结构的综合遗传测 度(SPC),绘制了综合遗传测度和主成分(PC)的“克立格”地图,分析了其群体遗传空间结构特性。与其他空间 插值或平滑方法相比,人类群体遗传空间结构的“克立格”模型具有明显优点:1)“克立格”估计以空间遗传变异函 数模型为基础,在绘制空间遗传结构地图之前,可利用变异函数模型定量分析所研究基因座(或多基因座)的空间 遗传异质性;2)“克立格”插值方法是真正意义上的无偏估计模型,它利用待估区域周围的已知群体遗传调查点数 据,并充分考虑调查点的空间影响范围,给出待估区域的最优估计值;3)“克立格”模型允许估计插值误差,这种插 值误差既可用于评价空间估计效果,又可通过绘制误差地图指导在误差过高的地点增加新的群体遗传调查样本 点,以优化估计效果。然而,人类群体遗传空间结构的“克立格”模型也存在一定缺点:1)若不能用任何理论遗传变 异函数模型拟合观察遗传变异函数?
By applying Krieger technology to the field of human population genetics, a “Krieger” model of human population’s genetic space structure is constructed, and its principle and calculation method are discussed. Taking the HLA A locus as an example, the “kriging” model was used to quantitatively analyze the spatial heterogeneity of HLA A loci in Chinese population. The principal component analysis of the spatial data matrix of HLA A gene frequencies was carried out, The population genetic structure of the integrated genetic measure (SPC), drawn a comprehensive genetic measure and the principal component (PC) of the “kriging” map, analyzed its population genetic structure characteristics. Compared with other methods of spatial interpolation or smoothing, the “Kriging” model of human population’s genetic space structure has obvious advantages: 1) The “Kriging” estimation is based on the spatial genetic variation function model and prior to drawing the spatial genetic structure map , The variability function model can be used to quantitatively analyze the spatial genetic heterogeneity of the studied locus (or locus); 2) The Kriging interpolation method is a truly unbiased estimation model that takes advantage of the region around the region to be estimated Of the known population genetic checkpoint data, and give full consideration to the spatial extent of the survey point, giving the best estimate of the region to be estimated; 3) “kriging” model allows estimation of interpolation error, this interpolation error can be used Evaluation of spatial estimation results, but also by mapping the error map guidance in the place of over-error in the new population of genetic survey sample points to optimize the estimation effect. However, the “kriging” model of the genetic structure of human population also has some shortcomings: 1) Can we observe the genetic variability function by any theoretical genetic variation function model?