摘要
花生含油量对单位面积产油量至关重要。该性状受多个微效基因控制,但可用的紧密连锁标记十分有限,传统的分子标记辅助选择育种准确性不高。全基因组选择作为一种新的育种方法,可实现对数量性状的早期预测;近红外光谱分析可对作物品质性状(如含油量等)进行无损检测。通过两者优势互补,建立花生含油量全基因组选择和近红外光谱筛选联合的育种技术,探讨影响花生含油量全基因组选择预测准确性的因素,为花生分子育种奠定理论基础。本研究以216个重组自交系为材料构建训练群体;分别以139、464和505株F_(2)、F_(3)和F_(4)为材料构建育种群体;利用自主开发的“PeanutGBTS40K”液相芯片进行基因分型,开展含油量全基因组选择育种模型分析;通过联合全基因组选择和近红外光谱筛选技术,开展花生含油量性状的育种应用,并评价其育种效果。结果显示,对训练群体进行基因分型后,总共获得30,355个高质量SNPs,并用于11个全基因组预测的模型选择分析。含油量预测准确性最高的模型为rrBLUP,其次是randomforest和svmrbf。以重组自交系为预测群体,F_(2)、F_(3)和F_(4)各世代含油量的预测准确性分别为0.116、0.128和0.119;以重组自交系叠加上一轮的育种群体为预测群体,各世代含油量的预测准确性分别为0.116、0.131和0.160。全基因组选择联合近红外筛选要比单独的全基因组选择对各世代的含油量选择效果提高1.8%、2.7%和3.4%;与单独的近红外筛选相比,差异不显著(0.10%、0.06%和0.07%);而近红外筛选与全基因组选择相比,含油量可显著提高1.7%、2.6%和3.3%。通过联合全基因组选择和近红外光谱筛选育种,F_(3)和F_(4)分别比F_(2)的含油量提高1.2%和1.0%。F_(4)总共获得16个入选改良株系,有10个株系含油量≥55.0%,其中2个株系(SF_(4)_201和SF_(4)_379)的理论产量分别比对照增产7.0%和11.1%。本研究通过建立花生含油量性状的全基因组选择-近红外光谱筛选联合育种技术,可有效实现花生含油量性状的遗传改良。
Oil content is a crucial trait for the yield of oil per unit area in peanut.This trait is controlled by multiple minor genes,and its avaliable tightly linked markers are very limited,resulting in low breeding accuracy in traditional molecular marker assisted selection.Genomic selection(GS),as a new breeding method,could achieve early prediction of quantitative traits.Near infrared ray(NIR)technology can non-destructively detect seed quality traits,such as oil content.By combining the advantages of the two breeding technologies,we have established a breeding technology that combined GS and NIR for breeding peanut oil content,and explored the factors that affected the accuracy of GS for peanut oil content.This study lays a theoretical foundation for peanut molecular breeding.Here,a total of 216 recombinant inbred lines were used as a training population.The F_(2)(139),F_(3)(464),and F_(4)(505)were used to construct the breeding populations.Genotyping was carried out using the self-developed“PeanutGBTS40K”liquid chip.The breeding application of oil content was conducted using a GS and NIR jointed breeding technology,and evaluated its breeding effects.The results showed that after genotyping the training population,a total of 30,355 high-quality SNPs were obtained,and used for 11 GS models selection analyses.The rrBLUP model showed the highest accuracy,followed by randomforest and svmrbf.The GS prediction accuracy of F_(2),F_(3),and F_(4) was 0.116,0.128,and 0.119,respectively,using recombinant inbred lines as the training population.Accordingly,the prediction accuracy was 0.116,0.131,and 0.160,respectively,using a superimposed training population.Compared with the GS,the GS-NIR can improve oil content by 1.8%,2.7%,and 3.4%for each generation.Compared with the NIR,there was no significant difference(0.1%,0.06%,and 0.07%).Compared with the GS,the NIR can significantly improve oil content by 1.7%,2.6%,and 3.3%for each generation.Through the combined technologies,compared to F_(2),the oil content of F_(3) and F_(4) increased by 1.2%and 1.0%,respectively.Finally,a total of 16 improved lines were obtained in F_(4),of which 10 lines had oil content≥55.0%.Among them,two lines(SF_(4)_201 and SF_(4)_379)had a theoretical yield increase of 7.0%and 11.1%,respectively,compared to the control variety.This study suggested that oil content could be effectively improved through GS combined with NIR in peanut.
作者
鲁清
刘浩
李海芬
王润风
黄璐
梁炫强
陈小平
洪彦彬
刘海燕
李少雄
LU Qing;LIU Hao;LI Hai-Fen;WANG Run-Feng;HUANG Lu;LIANG Xuan-Qiang;CHEN Xiao-Ping;HONG Yan-Bin;LIU Hai-Yan;LI Shao-Xiong(Crops Research Institute,Guangdong Academy of Agricultural Sciences/Guangdong Provincial Key Laboratory of Crop Genetic Improvement/South China Peanut Sub-Center of National Center of Oilseed Crops Improvement,Guangzhou 510640,Guangdong,China)
出处
《作物学报》
CAS
CSCD
北大核心
2024年第4期969-980,共12页
Acta Agronomica Sinica
基金
2022年省级乡村振兴战略专项资金种业振兴项目(2022-NPY-00-022)
广东省重点领域研发计划项目-现代种业(2020B020219003,2022B0202060004)
财政部和农业农村部国家现代农业产业技术体系建设专项(花生,CARS-13)
广东省农业科学院农业优势产业学科团队项目(202104TD)
广东省农业科学院协同创新中心项目(XTXM202203)资助。
关键词
花生
含油量
全基因组选择
近红外光谱分析
基因组育种值
peanut(Arachis hypogaea L.)
oil content
genomic selection
near infrared ray
genomic breeding value
