Context-guided diffusion for out-of-distribution molecular and protein design

Conference item

Generative models have the potential to accelerate key steps in the discovery of novel molecular therapeutics and materials. Diffusion models have recently emerged as a powerful approach, excelling at unconditional sample generation and, with data-driven guidance, conditional generation within their training domain. Reliably sampling from high-value regions beyond the training data, however, remains an open challenge'with current methods predominantly focusing on modifying the diffusion process itself. In this paper, we develop context-guided diffusion (CGD), a simple plug-and-play method that leverages unlabeled data and smoothness constraints to improve the out-of-distribution generalization of guided diffusion models. We demonstrate that this approach leads to substantial performance gains across various settings, including continuous, discrete, and graph-structured diffusion processes with applications across drug discovery, materials science, and protein design.

Authors: Klarner, L; Rudner, TGJ; Morris, GM; Deane, CM; Teh, YW

• Publication status: Published
• Peer review status: Peer reviewed
• Publisher: PMLR
• Host title: Proceedings of the 41st International Conference on Machine Learning
• Pages: 24770-24807
• Series: Proceedings of Machine Learning Research
• Series number: 235
• Publication date: 2024-08-01
• Event title: 41st International Conference on Machine Learning (ICML 2024)
• Event location: Vienna, Austria
• Event website: https://icml.cc/Conferences/2024
• Event start date: 2024-07-21
• Event end date: 2024-07-27
• EISSN: 2640-3498
• Language: English