Enhancing Wheat Crop Resilience with Generative Adversarial Networks

Common wheat, Triticum aestivum, is a staple crop that is facing problems such as drought, soil salinity and extreme temperature fluctuations, which are likely to increase with climate change. Although conventional breeding techniques are quite effective, they are slow and tedious and thus, there is a need for sophisticated computational techniques to create adaptive plant varieties. This paper aims to present how Generative Adversarial Networks (GANs) can be used to model and predict genetic diversity of Triticum aestivum, with regard to qualities like drought tolerance and salt adaptation. The research is based only on the DNA chromosome dataset of Triticum aestivum and the GANs are used to create synthetic climate adaptive genetic profiles. These profiles should reflect the actual genetic variation and help to find new ways of combining the traits for increased tolerance. This approach is a contribution to solving the limitations of the existing methods and, therefore, improves the efficiency of the crop development process, as well as the sustainability of agriculture and food production. The study introduces a GAN-based framework for realistic genetic simulations that are validated against experimental data. The GAN model successfully identifies and simulates crucial genetic variations, leading to enhanced trait prediction and selection for breeding. The results validate the GAN’s effectiveness with an accuracy of 92.31 percent in generating realistic, adaptive genetic profiles, which can be integrated into crop improvement programs.