Work

Genomic selection accuracy for the yield trait in alfalfa half-sib families based on GWAS strategies

Year: 2025

Type: article

Abstract: Alfalfa (Medicago sativa L.), a globally vital perennial forage supporting livestock, faces stagnant yield improvement due to polyploid complexity and slow genetic gains. To enhance genomic selection (GS) for fresh weight yield, this study integrated genome-wide association study (GWAS) strategies in 136 alfalfa half-sib families. Phenotypic data from five cuts collected during 2020–2022 were modeled into Best Linear Unbiased Predictions (BLUPs) to control environmental noise. GWAS, using these BLUPs, identified 19 significant single nucleotide polymorphisms (SNPs) associated with fresh weight per plant (FW). The prediction accuracy of 13 GS models (including gBLUP, rrBLUP, Bayesian, and machine learning methods) was evaluated. Strikingly, GWAS-informed SNP subsets boosted accuracy to 0.85–0.90. Moreover, the wgBLUP model, incorporating SNP-trait association strength as genomic weights, achieved high accuracy (0.90) even with all SNPs. These results demonstrate that leveraging GWAS data through SNP prioritization or weighted matrices significantly overcomes historical breeding bottlenecks and accelerates genetic gain, while this study proposes wgBLUP as an optimal strategy for accelerating alfalfa yield breeding in half-sib populations.

Source: Grass Research

DOI: 10.48130/grares-0025-0028