CROPGRO大豆花期模拟模型品种参数的遗传学解析  

Genetic analysis of cultivar parameters of CROPGRO soybean flowering simulation model

作　　者：许俊杰 闫文亮 赵团结[2] 姜海燕[1] XU Junjie;YAN Wenliang;ZHAO Tuanjie;JIANG Haiyan(College of Artificial Intelligence,Nanjing Agricultural University,Nanjing 210095,China;National Center for Soybean Improvement,Nanjing Agricultural University,Nanjing 210095,China)

机构地区：[1]南京农业大学人工智能学院,江苏南京210095 [2]南京农业大学国家大豆改良中心,江苏南京210095

出　　处：《南京农业大学学报》2021年第4期778-788,共11页Journal of Nanjing Agricultural University

基　　金：国家自然科学基金项目(31872847);江苏省现代作物生产协同创新中心项目(JCIC-MCP)。

摘　　要：[目的]DSSAT-CROPGRO大豆花期模拟模型(简称CROPGRO-Soybean-Flower模型)是量化开花时间与光周期关系的生理生态过程模型。本文旨在通过单核苷酸多态性(SNP)标记对CROPGRO-Soybean-Flower模型的品种参数(cultivar parameter,CP)进行遗传解析。[方法]以183个江淮大豆种质为试验材料,于2011—2018年获得南京、当涂、盐城等10个环境的花期数据和消除环境效应的花期最佳线性无偏预测值(best linear unbiased prediction,BLUP)。将开花时间性状分解为临界短日照时数(critical short day,CSDL)、光周期敏感性(photoperiod sensitivity,PPSEN)和出苗到开花时长(emergence to flowering time,EMFL)3个CROPGRO大豆花期模拟模型的品种参数。利用供试大豆材料的实测花期以及BLUP值,采用GLUE(generalized likelihood uncertainty estimation)方法对模型的品种参数进行校准,根据模拟结果选择合适的品种参数集。利用混合线性模型(mixed linear model,MLM_Q+K)、mrMLM(multi-locus random-SNP-effect mixed linear model)模型等7种全基因组关联分析方法,对花期BLUP值、CROPGRO-Soybean-Flower模型品种参数集与60712个SNP标记进行定位研究,在显著关联SNP位点上下游500 kb区间内筛选候选基因,并通过生物信息学网站和前人研究结果预测其功能。[结果]利用BLUP值校正CROPGRO-Soybean-Flower模型的品种参数,模型模拟结果的均方根误差的均值(average root mean square error,ARMSE)为1.65 d,比利用实测花期校正的ARMSE降低1.82 d。选用花期BLUP值校正的3个模型的品种参数集作为全基因组关联分析的数据集,检测到花期BLUP值、CSDL、PPSEN和EMFL的显著关联QTN(quantitative trait nucleotide)个数分别为6、10、5和1,所定位的SNP位点可解释0.424%~19.221%的CROPGRO-Soybean-Flower模型品种参数变异。被检测到的22个QTN中,有72.73%已被前人研究报道与大豆开花时间、光周期敏感性、生殖生长时期等性状有关。对候选基因功能预�[Objectives]DSSAT-CROPGRO soybean flowering simulation model(CROPGRO-Soybean-Flower model)is a physiological and ecological process model to quantify the relationship between flowering time and photoperiod.The purpose of our study was to operate a genetic analysis of cultivar parameter(CP)in CROPGRO-Soybean-Flower model via single nucleotide polymorphism(SNP)markers.[Methods]183 germplasms of Yangtze-Huai soybean breeding line were used as experimental materials.The flowering time and best linear unbiased prediction(BLUP)values were obtained under ten planting environments including Nanjing,Dangtu,Yancheng from years of 2011 to 2018.Three CP were decomposed from flower time,namely critical short daylight length(CSDL),photoperiod sensitivity(PPSEN),and emergence to flowering length(EMFL).These CP were calibrated via the generalized likelihood uncertainty estimation(GLUE)algorithm with BULP and measured flowering time data of tested soybean.Then,we selected the appropriate CP as the association analysis data set according to the simulation results of the flowering time.There were 7 genome-wide association analysis study methods,for example,the mixed liner model(MLM)and multi-locus random-SNP-effect mixed linear model(mrMLM).They were used to perform an association analysis on the appropriate CP,BLUP with 60712 SNP.The candidate genes were screened in the 500 kb interval of the significantly associated quantitative trait nucleotides(QTN)upstream and downstream,and their functions were predicted through bioinformatics websites and previous research results.[Results]The average root mean square error(ARMSE)of BLUP-based CROPGRO-Soybean-Flower model simulation results was 1.65 d,which reduced by 1.82 d compared with the simulation results caused by measured flowering data.Hence,for GWAS analysis,we chose BLUP and three calibrated cultivar parameter sets from them.The number of significant QTN detected to be associated with BLUP,CSDL,PPSEN,and EMFL were 6,10,5,and 1,respectively,and the locus obtained by the variety par

关 键 词：大豆 CROPGRO 品种参数 最佳线性无偏预测值 全基因组关联分析 

分 类 号：S565.1[农业科学—作物学]