From Molecules to Models: A Guide to AI Drug Discovery with Python

Developing a single drug takes 12 years with only a 12% chance of success. AI is changing this dramatically: the first AI designed drug have reached human trials, Alphafold won a Nobel Prize and pharmaceutical companies have committed billions to AI partnerships. The best part? The Python ecosystem you already know is powering this revolution. This talk introduces AI drug discovery to Python developers with no biology or chemistry background required. We'll start with the key insight that makes this field accessible: molecules are data structures, proteins are strings, and drug target binding is just API matching. Through a demo, attendees will see how to represent and visualize molecules with RDKit, convert chemical structures into ML ready vectors, predict drug properties like toxicity and solubility using graph neural networks with DeepChem and predict 3D protein structures in seconds using the ESMFold API.