作者: Jian-cheng Wang , Jin Hu , Xin-xian Huang , Sheng-chun Xu
关键词:
摘要: One hundred and sixty-eight genotypes of cotton from the same growing region were used as a germplasm group to study validity different genetic distances in constructing core subset. Mixed linear model approach was employed unbiasedly predict genotypic values 20 traits for eliminating environmental effect. Six commonly (Euclidean, standardized Euclidean, Mahalanobis, city block, cosine correlation distances) combining four hierarchical cluster methods (single distance, complete unweighted pair-group average Ward’s methods) least distance stepwise sampling (LDSS) method subsets. The analyses variance (ANOVA) evaluating parameters showed that validities inferior those Mahalanobis block distances. Standardized Euclidean slightly more effective than principal analysis validated course practical covariance matrix accessions might be ill-conditioned when calculate at low percentages, which led bias small-sized subset construction. is recommended construction with LDSS method.