作物学报2018，Vol.44Issue(1)：43-52,10.DOI:10.3724/SP.J.1006.2018.00043
群体构成方式对大豆百粒重全基因组选择预测准确度的影响
Effect of Population Structure on Prediction Accuracy of Soybean 100-Seed Weight by Genomic Selection
摘要
Abstract
Hundred-seed weight is an important yield component and has positive relationship with soybean yield under certain conditions. The genetic gain of 100-seed weight based on traditional breeding or markers assisted-selection is limited because it is controlled by plenty of small effect genes. Genomic selection offers an approach to accelerate the soybean 100-seed weight breeding. However, the effect of population structure on soybean 100-seed weight prediction accuracy has not been elaborated. In our study 280 soybean varieties with phenotypic data evaluated in multi-location in 2008–2012 and 5361 SNPs genotype were used to explore the effect of population structure on 100-seed weight prediction accuracy. The best linear unbiased prediction of 100-seed weight of each variety was calculated according to mixed linear model. Ridge regression best linear unbiased prediction and five-fold cross validation were used to estimate the 100-seed weight prediction accuracy. Our research showed that the range of 100-seed weight, which was from –0.15 to +0.75. Hundred-seed weight prediction accuracy was affected by population struc-ture significantly. The prediction accuracy within a subset (+0.24 to +0.75) was higher than that between subsets (-0.15 to +0.29). When the genetic distance between subsets increased from 0.1566 to 0.2201, the 100-seed weight prediction accuracy was de-creased by 27.87％. Compared with random sampling training population, the training population composed based on genetic structure improved 100-seed weight prediction accuracy by 2.34％. In summary, we are clear about the soybean 100-seed weight genomic selection accuracy and the effect of population structure on genomic selection accuracy. The genomic selection is an efficient method to improve the soybean breeding.关键词
大豆/百粒重/全基因组选择/预测准确度/遗传结构Key words
Glycine max/100-seed weight/genomic selection/prediction accuracy/genetic structure引用本文复制引用
马岩松,张万海,吴纪安,胡国华,栾晓燕,付亚书,郭泰,王曙明,韩天富,张孟臣,张磊,刘章雄,苑保军,郭勇,Jochen C.REIF,江勇,李文滨,王德春,邱丽娟,文自翔,魏淑红,杨春明,王会才,杨春燕,卢为国,徐冉..群体构成方式对大豆百粒重全基因组选择预测准确度的影响[J].作物学报,2018,44(1):43-52,10.基金项目
本研究由国家转基因生物新品种培育重大专项(2014ZX08004001)和中国农业科学院农业科技创新项目资助.This study was supported by the National Major Project for Developing New GM Crops (2014ZX08004001) and the Agricultural Science and Technology Innovation Program (ASTIP) of Chinese Academy of Agricultural Sciences. （2014ZX08004001）
