摘要
为提高育种家选种选育的效率,运用统计学手段对两组大豆的基因型数据及对应的真实性状设置不同遗传力及有效位点数,模拟不同性状进行运算来预测精准值。采用交叉验证重复百次使预测值趋于稳定,对比分析Bayesian B及常见的几种基因组选择方法。结果图形直观展示出Bayesian B方法效果较好且较稳定,说明Bayesian B方法在大豆全基因组选择上有较明显的优势,可以在大豆育种分析中辅助使用。同时,介绍了Bayesian B方法的应用,以便指导大豆品种选育。
In order to improve the breeding efficiency, different heritability and effective loci were set up by the genotype data and corresponding real characters of two groups of soybean to simulate the accuracy of different characters, statistical methods were used to simulate different characters to predict precise values. Cross validation repeated hundreds of times to stabilize the prediction value, compare and analyze Bayesian B and several common methods of genomic selection. The results showd that the Bayesian B method was better and more stable, indicating that the Bayesian B method had obvious advantages in the selec- tion of soybean whole genome, which could be used in soybean breeding. At the same time, the application of Bayesian B method was introduced to guide the soybean breeding.
作者
唐友
郑萍
王嘉博
张继成
TANG You;ZHENG Ping;WANG Jia-bo;ZHANG Ji-cheng(Electrical and Information Engineering College,Jilin Agricultural Seienee and Technology University, Jilin 132101, China;College of Electrical and Information,Northeast Agricultural University, Harbin 150030, China;Institute of Animal Husbandry,Heilon~iang Academy of Agrleultural Sciences, Harbin 150086, China)
出处
《大豆科学》
CAS
CSCD
北大核心
2018年第3期353-358,共6页
Soybean Science
