AI-ENGAGE: Harnessing AI for Precision Genomic Selection Models in Wheat

The project addresses limitations in traditional wheat breeding by integrating vast datasets from genome sequencing, weather stations, and high-throughput phenotyping (e.g., drone imaging). It utilizes feature-embedded neural networks to combine genetic markers, environmental factors, and plant growth measurements to predict wheat performance. A core set of 400 wheat lines will be genotyped and evaluated over three years in two locations under normal and drought conditions. Environmental data will be collected to create indices and response parameters. The neural networks will incorporate sequence variants, environmental indices, and quantitative trait loci identified via genome-wide association studies. The project also includes training initiatives for agricultural scientists.