论文部分内容阅读
以 992份水稻品种为材料 ,通过 13个性状的测定和评价 ,进行水稻核心种质的构建方法研究。采用调整无偏预测法 (AUP)无偏预测水稻性状的基因型值 ,用基因型值计算不同遗传材料间的马氏距离 ,分别用 3种聚类方法 (最长距离法、不加权类平均以及离差平方和 )结合随机取样法、优先取样和变异度取样 3种抽样方法构建出 9个不同的水稻核心种质。水稻核心种质和原种质资源群体的方差差异百分率、均值差异百分率、极差符合率和变异系数变化率的比较结果表明 ,采用不加权类平均法进行多次聚类、结合变异度取样的方法可以较好地构建出水稻核心种质。基于基因型值构建的水稻核心种质比利用表现型值构建的核心种质更能代表原种质资源群体的遗传多样性。
Taking 992 rice varieties as materials, the methods of constructing the core collection of rice were studied through the determination and evaluation of 13 traits. The unbiased predictive value (AUP) was used to predict the genotypes of rice traits. The Mahalanobis distances between different genetic materials were calculated using genotype values. Three kinds of clustering methods (longest distance method, non-weighted average And sum of squares of deviations) were used to construct nine different rice core collections by combining three sampling methods: random sampling, sampling with priority and sampling with variation degree. The variance variance percentage, mean difference percentage, poor coincidence rate and variation coefficient of rice germplasm and original germplasm resource population showed that the non-weighted average method was used for multiple clustering and the variation degree sampling The method can well construct the rice core collection. The core collection constructed based on the genotype value is more representative of the genetic diversity of the original germplasm resource population than the core collection constructed using phenotypic values.