GEG2P is a genetic algorithm-based ensemble learning method that improves genomic prediction accuracy by an average of 4.02% across major crops. By dynamically optimizing a combination of 20 diverse machine learning models, it enables more accurate genotype-to-phenotype prediction and provides deep insights into influential genetic markers using SHAP analysis. This robust tool supports intelligent breeding for enhanced crop varieties.
GEG2P is an advanced genetic algorithm-based ensemble learning method designed to enhance genotype-to-phenotype prediction in crop breeding. By integrating 20 distinct base learners—spanning statistical, machine learning, and deep learning approaches—the method dynamically optimizes combinations and weights to provide superior predictive accuracy. It addresses the challenge of inconsistent performance across different crops and complex traits, serving as a powerful tool for accelerating intelligent crop variety development and improving global food security.
Huazhong Agricultural University is a public research university in Wuhan with a strong focus on agriculture and the life sciences, coupling fundamental inquiry with field‑ready application. Industry engages through co‑located experimental farms and controlled‑environment facilities, shared cores for genomics and phenotyping, and biosafety‑equipped laboratories. Field stations and demonstration bases across the region enable multi‑season trials and scale‑up with growers and agri‑tech partners, while proximity to Wuhan’s high‑tech corridor and logistics hub streamlines collaboration. Research is supported by competitive national and provincial funding, including the National Natural Science Foundation of China and related ministries. A dedicated technology transfer office facilitates IP management, licensing, and enterprise partnerships.