Bridging the Gap between Learning and Inference for Diffusion-Based Molecule Generation
This work addresses a specific bottleneck in molecular generation for drug discovery, offering an incremental improvement over existing diffusion methods.
The paper tackles the exposure bias problem in diffusion-based molecule generation by proposing GapDiff, a training framework that uses model-predicted conformations as probabilistic ground truth to reduce distributional disparities, resulting in improved affinity scores on the CrossDocked2020 dataset.
The efficacy of diffusion models in generating a spectrum of data modalities, including images, text, and videos, has spurred inquiries into their utility in molecular generation, yielding significant advancements in the field. However, the molecular generation process with diffusion models involves multiple autoregressive steps over a finite time horizon, leading to exposure bias issues inherently. To address the exposure bias issue, we propose a training framework named GapDiff. The core idea of GapDiff is to utilize model-predicted conformations as ground truth probabilistically during training, aiming to mitigate the data distributional disparity between training and inference, thereby enhancing the affinity of generated molecules. We conduct experiments using a 3D molecular generation model on the CrossDocked2020 dataset, and the vina energy and diversity demonstrate the potency of our framework with superior affinity. GapDiff is available at \url{https://github.com/HUGHNew/gapdiff}.