GraphGUIDE: interpretable and controllable conditional graph generation with discrete Bernoulli diffusion
This work addresses the challenge of generating graphs, such as drug-like molecules, with specific desired properties in an interpretable way, which is incremental over existing diffusion methods for graphs.
The authors tackled the problem of interpretable and controllable conditional graph generation by proposing GraphGUIDE, a diffusion-based framework that enables full control over arbitrary structural properties without predefined labels, achieving state-of-the-art performance on several graph datasets.
Diffusion models achieve state-of-the-art performance in generating realistic objects and have been successfully applied to images, text, and videos. Recent work has shown that diffusion can also be defined on graphs, including graph representations of drug-like molecules. Unfortunately, it remains difficult to perform conditional generation on graphs in a way which is interpretable and controllable. In this work, we propose GraphGUIDE, a novel framework for graph generation using diffusion models, where edges in the graph are flipped or set at each discrete time step. We demonstrate GraphGUIDE on several graph datasets, and show that it enables full control over the conditional generation of arbitrary structural properties without relying on predefined labels. Our framework for graph diffusion can have a large impact on the interpretable conditional generation of graphs, including the generation of drug-like molecules with desired properties in a way which is informed by experimental evidence.