Dr. Mohsen Yoosefzadeh Najafabadi
Research Overview
Dr. Mohsen Yoosefzadeh Najafabadi’s research lies at the intersection of plant breeding, computational biology, and artificial intelligence. His work integrates genomics, multiomics, and AI methods to develop dry bean varieties that are climate-resilient, high yielding, and high in quality.
A key focus of his research is the development of AI-powered tools to support plant breeders. He created BeanGPT, a generative AI platform that assists breeders in selecting suitable varietals for specific field conditions. The platform functions as a co-pilot for plant scientists and is built on extensive scientific literature and decades of accumulated domain knowledge.
Research Keywords
Plant breeding and genetics | Genomics and multiomics | Artificial Intelligence in Agriculture | Generative AI for decision support | Climate-resilient crop development | Computational biology
Dr. Mohsen's Projects
BeanGPT
AI-Enabled Multi-Omics Mapping of Seed Coat Colour Stability in Dry Beans
This project applies AI-assisted data integration and pattern discovery to understand seed coat colour stability in cranberry and pinto beans. Machine learning models are used to jointly analyze genomic variation, DNA methylation profiles, hyperspectral reflectance data, and temporal storage measurements. AI methods identify nonlinear relationships and subtle interactions that are difficult to detect using traditional statistics alone. The outcomes include AI-informed genetic and epigenetic markers that breeders can use to predict colour stability early in the breeding cycle.
AI-Assisted Anthracnose Resistance Screening and Prediction
To address increasing anthracnose pressure, this project combines AI-based image analysis, genomics, and field phenotyping to improve resistance screening. Deep learning models analyze drone and ground-based imagery to detect early disease symptoms and quantify severity across large populations. These phenotypic outputs are integrated with molecular marker data using machine learning to predict resistance levels and prioritize lines for advancement. AI enables faster, more consistent screening while reducing reliance on subjective visual scoring.
AI-Driven Early Prediction of Canning Quality in Dry Beans
This project introduces AI-powered image analysis and predictive modeling to assess canning quality traits at earlier breeding stages. Computer vision models extract quantitative features from cooked and canned beans, while machine learning models link these features to historical quality outcomes and genetic backgrounds. By integrating AI into the evaluation pipeline, breeders can simultaneously select for yield and processing quality, reducing late-stage failures and improving alignment with processor requirements.
AI-Based High-Throughput Field Phenotyping and Stress Response Analysis
This project leverages AI-enabled remote sensing and environmental modeling to quantify crop performance under variable field conditions. Machine learning models analyze drone imagery, weather data, and soil information to capture growth dynamics, stress responses, and yield stability. AI is used to model genotype–environment interactions and identify lines that maintain performance across unpredictable climates, supporting climate-resilient breeding strategies.
AI Integration of Historical Trial Data for Predictive Breeding
Decades of dry bean trial data are harmonized and analyzed using AI-based data curation, feature extraction, and predictive modeling. Machine learning algorithms uncover long-term trends, trait correlations, and stability patterns that are not apparent from individual trials. These insights feed directly into AI tools such as BeanGPT, enabling data-informed breeding decisions that build on past knowledge rather than rediscovering it each cycle.
Partner Organizations
- Ontario Bean Growers
- Hensall Co-Op
- Agilent Technology
- Sprague Food
- OMAFA (Ontario Ministry of Agriculture, Food and Agribusiness)
- Agricultural Research and Innovation Ontario (ARIO)
- Natural Sciences and Engineering Research Council of Canada (NSERC)
- AgData Consortium
- Research Innovation Office (RIO) at University of Guelph