Titolo della tesi: Computational and machine learning approaches for the rational design and discovery of novel bioactive compounds
The use of computational approaches in drug design and discovery has become an essential tool for identifying, prioritising and optimising biologically active compounds in the last decades. These approaches are a valuable integral part of the preliminary stages of the drug discovery pipeline, helping to expedite the drug development process in a more cost-efficient way. They significantly reduce the time and resources required by conventional experimental strategies, such as chemical synthesis and biological testing, while covering a wider chemical space. The most recent impact of big data handling and artificial intelligence (AI) in the field holds great promise to revolutionise and streamline the entire drug discovery process. The extensive variety of computational tools are broadly classified as ligand-based or structure-based methods, depending on the availability of high-resolution structural data about the target. A combination of these methods often produces the most successful stories. This PhD thesis presents the application of computational approaches, chemometric and machine learning techniques to drug design and discovery projects. These techniques provide valuable support in unravelling information from data, assisting in ligand optimization and clarifying potential mechanisms of action. The combination of computational methods and advanced data analysis offers a comprehensive framework for advancing drug discovery, providing valuable support in navigating the complexities of modern pharmaceutical research.