高粱千粒重全基因组关联分析和候选基因预测

千粒重是作物产量构成的三要素之一。增加粒重是提高作物产量的重要途径。为阐明高梁千粒重的遗传机理,本试验以主要来自于我国16个高粱种植省份的242份高粱品种(系)组成的关联群体为研究材料,借助覆盖全基因组的2 015 850个高质量SNPs标记,采用多位点分析软件mrMLM 4.0 R软件包中的mrMLM、FASTmrMLM、FASTmrEMMA、ISIS EM-BLASSO、pLARmEB、pKWmEB 6种模型,对2018—2020年贵州贵阳、浙江杭州、海南乐东、海南陵水3年4点7个环境的高梁千粒重进行全基因组关联分析。结果表明,7个环境的高梁千粒重均表现连续正态分布,呈现明显的数量性状遗传特性,变异范围于10~50 g之间。利用6个模型共检测到323个与千粒重显著关联的数量性状核苷酸位点(QTNs),这些位点不均匀地分布在10条染色体上。单个QTN可解释0.4%~26.6%的表型变异。不同模型检测到的位点数目不同,FASTmrMLM模型最多,然后依次是pKWmEB模型、pLARmEB模型、mrMLM模型、ISIS EM-BLASSO模型和FASTmrEMMA模型。合并共同QTNs后得到96个显著影响千粒重的QTNs,其中4个QTNs与前人报道的高粱基因位点重叠,另有4个QTNs包含与水稻中克隆的粒重相关基因qGW7/GL7、BG2、OsARF4、RSR1、TGW6等同源的基因。本研究结果为探究高粱千粒重性状分子遗传机理和高粱分子设计育种提供了科学依据。

Genome-wide Association Analysis of Thousand Grain Weight and Candidate Gene Prediction in a Sorghum Sequenced Association Panel

Thousand grain weight (TGW) is one of the three important components of crop yield. Increasing grain weight is an effective way in the improvement of crop yield. In this study, to elucidate the genetic machanism of 1000-grain weigth an association mapping panel consisting of 242 sorghum landraces/cultivars collected mainly from 16 sorghum planting provinces across China was genotyped using whole-genome re-sequencing method. TGW of each accession was phenotyped across seven environments over three years from 2018 to 2020. Multi-locus association analyses based on 2 015 850 SNPs were carried out using mrMLM 4.0 R package which comprises the mrMLM, FASTmrMLM, FASTmrEMMA, ISIS EM-BLASSO, pLARmEB, and pKWmEB models. Results showed that TGW displayed a continuous and normal distribution as a typical quantitative trait. The variations of TGW were ranged from 10~50 g under seven environments. A total of 323 quantitative trait nucleotides (QTNs) for TGW were detected using six models. The phenotypic variation explained by individual QTN varied from 0.4% to 26.6%. Six multi-locus models revealed differential QTN numbers associated with TGW. The maximum number of associated QTNs was observed in FASTmrMLM model followed by pKWmEB model, pLARmEB model, mrMLM model, ISIS EM-BLASSO model, and FASTmrEMMA model. After merging the same QTNs, 96 consistently reliable QTNs were detected using multi-locus GWAS methods and/or at least two seasons, which were distributed unevenly on 10 chromosomes. Among them, 4 QTNs overlapped with previously reported QTLs. Besides, 5 candidate orthologous to documented rice grain weight genes, like GW7/GL7, BG2, OsARF4, RSR1, TGW6, were found to be located within four QTNs. These results help us to understand the genetic architecture of grain size and pave the way for exploration of underlying molecular mechanisms and molecular design breeding of this trait in sorghum breeding practices.