Digital platforms and computational chemistry methods: linking agri-food data to drug discovery and structural biology

The objective of this thesis is to apply advanced computational methods to support innovation across different but connected research areas, from sustainable agri-food systems to molecular studies related to human health. The work combines data-driven approaches with molecular modeling techniques to explore how digital tools can transform both agricultural and biomedical research. Within this framework, computational methods such as molecular docking, molecular dynamics, and free energy calculations (FEP) were used to study the behavior and binding properties of specific protein targets, including the serine/threonine kinase NEK6 and the receptor CD44. These studies aimed to understand molecular interactions, evaluate the stability of protein–ligand complexes, and identify new compounds with potential pharmacological relevance. Deep learning and machine learning tools were also employed to predict and refine the structure of CD44, providing insights into its conformational dynamics and its interaction with natural molecules such as hyaluronic acid. At the same time, this work was developed within the Agritech National Research Center, particularly in the activities of Spoke 9, led by the University of Siena. This research line focuses on improving the technological and qualitative characteristics of agri-food products and on identifying bioactive compounds from fruits, vegetables, and agricultural by-products. These natural molecules, many of which show antioxidant or anticancer properties, were integrated into computational workflows for virtual screening and molecular modeling. This allowed their potential biological activity to be studied through the same computational strategies used for traditional drug discovery. By linking the identification of bioactive compounds from the agri-food sector with molecular simulations and predictive modeling, this thesis demonstrates how digital and computational tools can act as a bridge between Agritech and biomedical research. The integration of data analytics, artificial intelligence, and molecular modeling supports both the valorization of natural products and the search for new therapeutic agents. In this way, the work highlights the shared scientific ground between sustainable agriculture and pharmaceutical innovation, showing how computational methods can generate value and knowledge across different domains of research.