Answers are grounded in this paper and cite their sources.
Soybean is a globally important economic and food crop. Plant height (PH) and flowering time (FT) are key agronomic traits that influence both yield formation and environmental adaptability. Genomic selection (GS) has become an essential tool for efficient soybean breeding. However, the limited characterization and utilization of epistatic interactions among non-allelic loci have constrained further improvements in GS prediction accuracy and restricted breeding efficiency. In this study, we conducted genome-wide epistasis detection for PH and FT in soybean and integrated the identified epistatic interaction pairs into six GS models. The results showed that incorporating epistatic effects significantly improved prediction accuracy across all models for both traits, with accuracy improvements ranging from 15.48% to 17.63% for PH and from 8.46% to 10.50% for FT. Further validation using transcription factor (TF)-target gene regulatory relationship analysis and AlphaFold2-based protein structure prediction confirmed the biological reliability and structural feasibility of these interaction pairs. Network construction and functional enrichment analyses revealed that the associated genes were significantly clustered within specific biological regulatory networks, in which optimal interaction pairs could coordinately regulate phenotypes through non-additive interactions. Collectively, these findings demonstrate that the systematic integration of epistatic interactions in soybean breeding can effectively enhance the predictive ability and selection efficiency of GS, providing a novel strategy for the efficient genetic improvement of complex traits.
No takes yet. Share an insight, caveat, or question.
Zhang et al. (2026) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: