This masterclass offers an engaging introduction to TorchDrug, a cutting-edge machine learning platform designed for drug discovery. It begins with an overview of the platform's purpose and structure, followed by a hands-on tutorial exploring data structures and graph operations within TorchDrug. Attendees will see practical demonstrations of pre-training, fine-tuning, and molecule generation. For instance, they'll learn how to manipulate molecular data structures with functions similar to PyTorch tensor operations and leverage pre-trained models for tasks like property prediction. Overall, the goal is to streamline drug discovery research by providing a user-friendly and efficient platform, saving researchers from the arduous task of building models from scratch.
Sign in to continue reading, translating and more.
Continue