MEG MODELS: A HOLISTIC, SYSTEMS-BASED MODELING TECHNIQUE FOR IMPROVED AGRICULTURAL PRODUCTION SYSTEM PERFORMANCE AND REDUCED POSTHARVEST LOSS

The overall goal of this project is to use novel modeling techniques to improve the performance and sustainability of crop production systems, particularly through the more accurate selection of the best-suited cultivars, an important cultural control, and the prediction of disease severity in untested environments through the use of weather and climatic indicators. The specific objectives of this project include:Evaluate 432 selected inbreds from the Ames panel and 108 F1 hybrids derived from this inbred panel for their resistance to Gibberella ear rot and mycotoxin contamination across multiple locations and years. Collect weather data for each environment.Determine the metabolic profiles of the selected Ames inbreds and a subset of their hybrids.Use machine learning techniques to extract information from environmental, phytochemical, and genomic data to predict (a) which entries will be best suited for specific locations and (b) which environments, based on weather predictors, will be at a greater risk of Gibberella ear rot disease pressure and mycotoxin contamination.