Table_1_Genome-wide analyses using multi-locus models revealed marker-trait associations for major agronomic traits in Sorghum bicolor.xlsx

Globally, sorghum is the fifth most important cereal crop, and it is a major crop in Ethiopia, where it has a high genetic diversity. The country’s sorghum gene pool contributes significantly to sorghum improvement worldwide. This study aimed to identify genomic regions and candidate genes associated with major agronomic traits in sorghum by using its genetic resources in Ethiopia for a genome-wide association study (GWAS). Phenotypic data of days to flowering (DTF), plant height (PH), panicle length (PALH), panicle width (PAWD), panicle weight (PAWT), and grain yield (GY) were collected from a GWAS panel comprising 324 sorghum accessions grown in three environments. SeqSNP, a targeted genotyping method, was used to genotype the panel using 5,000 gene-based single nucleotide polymorphism (SNP) markers. For marker-trait association (MTA) analyses, fixed and random model circulating probability unification (FarmCPU), and Bayesian-information and linkage-disequilibrium iteratively nested keyway (BLINK) models were used. In all traits, high phenotypic variation was observed, with broad-sense heritability ranging from 0.32 (for GY) to 0.90 (for PALH). A population structure, principal component analysis, and kinship analysis revealed that the accessions could be divided into two groups. In total, 54 MTAs were identified, 11 of which were detected by both BLINK and farmCPU. MTAs identified for each trait ranged from five (PAWT and GY) to fourteen (PH) representing both novel and previously identified quantitative trait loci (QTLs). Three SNPs were associated with more than one trait, including a SNP within the Sobic.004G189200 gene that was associated with PH and PAWT. Major effect SNP loci, Sbi2393610 (PVE = 23.3%), Sbi10438246 (PVE = 35.2%), Sbi17789352 (PVE = 11.9%) and Sbi30169733 (PVE = 18.9%) on chromosomes 1, 3, 5 and 9 that showed strong association signals for PAWD, DTF, GY and PALH, respectively, were major findings of this study. The SNP markers and candidate genes identified in this study provide insights into the genetic control of grain yield and related agronomic traits, and once validated, the markers could be used in genomics-led breeding.全球范围内，高粱是第五大重要谷类作物，而在遗传多样性丰富的埃塞俄比亚，高粱是当地主栽作物。该国的高粱基因库对全球高粱遗传改良具有重要贡献。本研究旨在利用埃塞俄比亚的高粱遗传资源开展全基因组关联分析（Genome-Wide Association Study, GWAS），挖掘与高粱主要农艺性状相关的基因组区域及候选基因。 本研究的试验群体包含324份高粱种质资源，在3种环境下种植，采集了开花天数（Days to Flowering, DTF）、株高（Plant Height, PH）、穗长（Panicle Length, PALH）、穗宽（Panicle Width, PAWD）、穗重（Panicle Weight, PAWT）及籽粒产量（Grain Yield, GY）的表型数据。采用靶向基因分型方法SeqSNP，基于5000个基因相关单核苷酸多态性（Single Nucleotide Polymorphism, SNP）标记对该群体进行基因分型。 在标记-性状关联（Marker-Trait Association, MTA）分析中，本研究采用了固定和随机模型循环概率统一（Fixed and Random model Circulating Probability Unification, FarmCPU）模型与贝叶斯信息与连锁不平衡迭代嵌套关键（Bayesian-Information and Linkage-Disequilibrium Iteratively Nested Keyway, BLINK）模型。 所有供试性状均表现出较高的表型变异，广义遗传力范围为0.32（籽粒产量）至0.90（穗长）。群体结构分析、主成分分析（Principal Component Analysis, PCA）及亲缘关系分析结果显示，供试种质可划分为两个类群。 本研究共鉴定到54个标记-性状关联位点，其中11个位点同时被BLINK与FarmCPU模型检测到。各性状对应的标记-性状关联位点数量范围为5（穗重与籽粒产量）至14（株高），涵盖新鉴定与已报道的数量性状位点（Quantitative Trait Loci, QTLs）。有3个SNP标记与多个性状相关联，其中位于Sobic.004G189200基因内的SNP同时与株高和穗重显著关联。 本研究的核心发现为位于1、3、5、9号染色体上的4个主效SNP位点：Sbi2393610（表型变异解释率Phenotypic Variance Explained, PVE = 23.3%）、Sbi10438246（PVE = 35.2%）、Sbi17789352（PVE = 11.9%）及Sbi30169733（PVE = 18.9%），分别与穗宽、开花天数、籽粒产量及穗长存在强关联信号。 本研究鉴定到的SNP标记与候选基因可为解析籽粒产量及相关农艺性状的遗传调控机制提供新的研究视角，待验证后的标记可应用于基因组主导育种。