HiEGP: An environment-informed genomic prediction framework that facilitates soybean genebank germplasm evaluation and climate-smart breeding.

HiEGP predicts soybean breeding traits from 1,421 germplasm accessions with pooled trait-wise correlations of 0.871 in observed environments and 0.775 in unobserved environments.

• Why it matters: Gene bank diversity is underused because phenotyping is costly and low-throughput, especially for photoperiod- and temperature-sensitive traits in region-specific settings.
• What they did: The framework combines feature-engineered hierarchical G×E modeling with environmental main effects, genotype-specific responses, genomic background, and environmental similarity using multi-environment phenotypes, genotypes, and bioclimatic/soil covariates.
• The result: HiEGP also supports environment-response GWAS, recovering maturity loci and finding 175 candidate regions, with mutant phenotypes supporting prioritized candidate genes and enabling regional adaptation assessment.

AI summary of the abstract · may contain errors · not medical advice · read the paper.