摘要： 用方差-协方差分析估计某一样本两个性状的遗传相关时，经常会遇到相关系数大于“1”的情况，从而造成遗传相关阵是非正定的。研究查明，这是由于基因型与环境的互作过大所造成的偏倚。作者将G×E剖分为遗传的和环境的两个部分，修改了常规的统计模型，可以避免遗传相关系数大于“1”的情况。
关键词:
遗传相关,
基因型与环境的互作
Abstract: The genotype by environment interaction may be biasing the genetic correlation coefficient estimates.If the interaction is big,a way described in the paper can be used to part the interaction variation(G×E)into different sources of linear(VGIX,VGIY,COVGI)and remender(VEIX,VEIY,COVED).rg=(COVG+COVGI)/[(VGX+VGIX)×(VGY+VGIY)]1/2Thus,we can avoid the issue of rgs exceeding 1.0.
Key words:
Genetic correlation,
Genotype by environment interaction
兰巨生. 避免遗传相关系数大于“ 1”的统计模型[J]. 华北农学报, 1987, 2(4): 28-31. doi: 10.3321/j.issn:1000-7091.1987.04.005.
Lan Jusheng. THE MODELS FOR AVOIDING ISSUE IN GENETIC CORRELATION COEFFICIENT EXCEEDING 1.0[J]. ACTA AGRICULTURAE BOREALI-SINICA, 1987, 2(4): 28-31. doi: 10.3321/j.issn:1000-7091.1987.04.005.