Enyew Muluken, Geleta Mulatu, Tesfaye Kassahun, Seyoum Amare, Feyissa Tileye, Alemu Admas, Hammenhag Cecilia, Carlsson Anders S
Department of Plant Breeding, Swedish University of Agricultural Sciences, Alnarp, Sweden.
Institute of Biotechnology, Addis Ababa University, Addis Ababa, Ethiopia.
BMC Plant Biol. 2025 Jan 17;25(1):69. doi: 10.1186/s12870-025-06077-w.
Root system architecture (RSA) plays an important role in plant adaptation to drought stress. However, the genetic basis of RSA in sorghum has not been adequately elucidated. This study aimed to investigate the genetic bases of RSA traits through genome-wide association studies (GWAS) and determine genomic prediction (GP) accuracy in sorghum landraces at the seedling stage. Phenotypic data for nodal root angle (NRA), number of nodal roots (NNR), nodal root length (NRL), fresh shoot weight (FSW), dry shoot weight (DSW), and leaf area (LA) were collected from 160 sorghum accessions grown in soil-based rhizotrons. The sorghum panel was genotyped with 5,000 single nucleotide polymorphism (SNP) markers for use in the current GWAS and GP studies. A multi-locus model, Fixed and random model Circulating Probability Unification (FarmCPU), was applied for GWAS analysis. For GP, ridge-regression best linear unbiased prediction (RR-BLUP) and five different Bayesian models were applied. A total of 17 SNP loci significantly associated with the studied traits were identified, of which nine are novel loci. Among the traits, the highest number of significant marker-trait associations (MTAs) was identified for nodal root angle on chromosomes 1, 3, 6, and 7. The SNP loci that explain the highest proportion of phenotypic variance (PVE) include sbi32853830 (PVE = 18.2%), sbi29954292 (PVE = 18.1%), sbi24668980 (PVE = 10.8%), sbi3022983 (PVE = 7%), sbi29897704 (PVE = 6.4%) and sbi29897694 (PVE = 5.3%) for the traits NNR, LA, SDW, NRA, NRL and SFW, respectively. The genomic prediction accuracy estimated for the studied traits using five Bayesian models ranged from 0.30 to 0.63 while it ranged from 0.35 to 0.60 when the RR-BLUP model was used. The observed moderate to high prediction accuracy for each trait suggests that genomic selection could be a feasible approach to sorghum RSA-targeted selection and breeding. Overall, the present study provides insights into the genetic bases of RSA and offers an opportunity to speed up breeding for drought-tolerant sorghum varieties.
根系结构（RSA）在植物适应干旱胁迫中起着重要作用。然而，高粱中RSA的遗传基础尚未得到充分阐明。本研究旨在通过全基因组关联研究（GWAS）探究RSA性状的遗传基础，并确定高粱地方品种在幼苗期的基因组预测（GP）准确性。从种植在土培根箱中的160份高粱种质中收集了节根角度（NRA）、节根数量（NNR）、节根长度（NRL）、地上部鲜重（FSW）、地上部干重（DSW）和叶面积（LA）的表型数据。该高粱群体用5000个单核苷酸多态性（SNP）标记进行基因分型，用于当前的GWAS和GP研究。采用多位点模型固定和随机模型循环概率统一法（FarmCPU）进行GWAS分析。对于GP，应用了岭回归最佳线性无偏预测（RR-BLUP）和五种不同的贝叶斯模型。共鉴定出17个与所研究性状显著相关的SNP位点，其中9个是新位点。在这些性状中，在第1、3、6和7号染色体上的节根角度鉴定出的显著标记-性状关联（MTA）数量最多。解释表型变异（PVE）比例最高的SNP位点包括：对于NNR、LA、SDW、NRA、NRL和SFW性状，分别为sbi32853830（PVE = 18.2%）、sbi29954292（PVE = 18.1%）、sbi24668980（PVE = 10.8%）、sbi3022983（PVE = 7%）、sbi29897704（PVE = 6.4%）和sbi29897694（PVE = 5.3%）。使用五种贝叶斯模型对所研究性状估计的基因组预测准确性在0.30至0.63之间，而使用RR-BLUP模型时在0.35至0.60之间。观察到的每个性状的中等至高预测准确性表明，基因组选择可能是一种针对高粱RSA进行定向选择和育种的可行方法。总体而言，本研究为RSA的遗传基础提供了见解，并为加速耐旱高粱品种的育种提供了机会。
