Overview
Explore the intersection of machine learning and molecular science in this 57-minute talk by Gianni De Fabritiis. Dive into deep learning techniques for molecular structures, generative models, and variational autoencoders like LigDream. Discover diffusion models and learn about PlayMolecules for drug discovery. Gain insights into TorchMD for machine learning in chemistry, including a deep dive into its architecture and the TorchMD-Net framework. Examine applications in protein science and conclude with a Q&A session. Access additional resources like datamol.io for molecular processing and join the M2D2 community for ongoing discussions in this rapidly evolving field.
Syllabus
- Intro
- Deep Learning Molecular Structures
- Generative Models of Structures
- LigDream: Variational Autoencoders
- Diffusion Models
- PlayMolecules: Acting on Drugs to Discover
- TorchMD: Machine Learning Chemistry
- TorchMD Deep Dive
- TorchMD-Net
- Machine Learning Proteins
- Conclusions
- Q+A
Taught by
Valence Labs