Genomic, phenomic, and enviromic data integration for bean yield prediction

Overview

This research integrates genomic, phenomic, and enviromic data to predict common bean behavior and yield performance across diverse environments and stress conditions. By combining whole-genome sequencing (WGS), high-throughput phenotyping platforms (HTPP), and environmental modeling, the approach enables more accurate prediction of complex quantitative traits such as yield and tolerance to abiotic stresses like phosphorus deficiency. The platform supports improved genotype selection, optimized fertilizer recommendations, and more reliable assessment of crop performance across agroecological zones.

Technical specifications

Key capabilities:

• Evaluation of 42 common bean genotypes across three locations for altitude adaptation and yield performance
• Whole-genome sequencing using the common bean reference genome assembly
• Assessment of 8 genotypes under stress and non-stress conditions (P2O5 deficiencies) measuring agronomic and physiological traits
• Spectral signature development combining agronomic, physiological, and environmental data
• Statistical and machine learning strategies including GWAS analysis, Genomic Selection (GS), linear mixed models, GBLUP, and deep learning approaches
• Multi and hyperspectral camera integration with vegetal and spectral indices for field-based phenotyping

Technology readiness level

The platform has completed initial validation with multi-location field trials and WGS analysis across 42 genotypes. A subset of 8 genotypes has been characterized under controlled stress conditions. Future validation will establish new field trials to continue data collection and refine prediction models using additional devices and expanded genomic tools. The technology is advancing toward broader deployment for biotechnology tool platforms supporting stress tolerance prediction.