AI Center Seminar - AI for Science series - Prof. Wengong Jin

The talk is jointly organized by the EPFL AI Center, the LTS2, and the LPDI labs as part of the AI for Science seminar series.

Hosting professors: Profs. Pierre Vandergheynst and Bruno Correia

Title
Generative AI for biomolecular design

Abstract
The discovery of biomolecules with desired properties is critical to advances in health, energy, and sustainability. This problem is challenging due to the combinatorial search space of biomolecules. In this talk, I will present how generative AI can be used to accelerate the discovery process across small molecules, proteins, and RNAs. First, I will introduce junction tree VAE (JT-VAE), a hierarchical generative model for small molecule design. The model successfully designs dye-like molecules with desired absorption wavelength, lipophilicity, and photooxidative stability. Second, I will introduce pTMEnergy, a differentiable statistical energy function derived from AlphaFold’s confidence prediction module. We incorporate pTMEnergy into BindEnergyCraft, a design pipeline that maintains the same optimization framework as BindCraft but replaces ipTM with our energy-based objective. BindEnergyCraft achieves higher in silico binder success rate than ipTM scores for both mini-proteins and RNA aptamers. In summary, our in silico and experimental results highlight the potential of generative AI for accelerating molecular discovery.

Bio
Wengong Jin is an assistant professor at Khoury College of Computer Sciences at Northeastern University. He is also a visiting research scientist in the Eric and Wendy Schmidt Center at Broad Institute. He obtained his PhD at MIT CSAIL, advised by Prof. Regina Barzilay and Prof. Tommi Jaakkola. His research focuses on geometric and generative AI models for drug discovery and biology. His work has been published in journals including ICML, NeurIPS, ICLR, Nature, Science, Cell, and PNAS, and covered by such outlets as the Guardian, BBC News, CBS Boston, and the Financial Times. He is the recipient of the BroadIgnite Award, Dimitris N. Chorafas Prize, and MIT EECS Outstanding Thesis Award.