Hit identification is a crucial step in the drug discovery process, where potential drug candidates are identified from a large chemical space. There are several methods for hit identification, including structure and ligand based-virtual screening, fragment-based drug design, and AI-driven drug design approaches.

In the context of de novo drug design, both generative methods and NLP are used to extract the information about existing data points, generate and optimize molecular structures based on desired properties. These algorithms can be trained on a variety of data sources, including molecular data, its biological profile, leading to a more informed hit generation process. In this talk, I will share an overview of these methods, practices, limitations, and case studies.