Genome-wide association and genomic prediction of seed nutritional traits and seed size in tepary bean (Phaseolus acutifolius) for low-input breeding

Tepary bean ( Phaseolus acutifolius A. Gray) is a heat- and drought-adapted legume with potential for low-input production but breeding resources for seed nutritional quality remain limited. A diversity panel of 206 accessions and four commercial checks was evaluated under certified organic field conditions in Uvalde, Texas, in 2024 for hundred-seed weight, seed width, total soluble seed protein, and nineteen free amino acids, and was genotyped with 49,384 high-quality SNPs generated by genotyping-by-sequencing. Genome-wide association analyses identified significant loci for protein percentage, seed size traits, and fifteen amino acid traits, with limited overlap among nutritional traits, indicating a largely modular genetic architecture. Essential amino acids of breeding interest, especially threonine, methionine, and lysine, showed distinct associations, while candidate genes near lead SNPs included a WNK kinase and 2-oxoglutarate/Fe(II)-dependent oxygenase for protein percentage, a vacuolar H + -ATPase-related region for hundred-seed weight, and an Aux/IAA-related region for seed width. Genomic prediction showed high predictive ability for seed size traits (r ≈ 0.90–0.96) and moderate predictive ability for protein and amino acid traits (r ≈ 0.15–0.45), supporting complementary use of marker-assisted and genomic selection. These results provide the first genome-scale breeding benchmarks for seed nutritional traits in tepary bean under a low-input production environment. Because the experiment was conducted in a single environment and season, treat the identified loci as environment-specific hypotheses that require validation across years and locations before routine deployment in breeding.