Mitigating Exposure Bias in Score-Based Generation of Molecular Conformations

TL;DR

This paper identifies and measures exposure bias in score-based generative models for molecular conformations, introduces a compensation method called Input Perturbation, and achieves state-of-the-art results on key datasets.

Contribution

It is the first to measure exposure bias in SGMs for molecular conformations and proposes a novel compensation algorithm that improves model accuracy and diversity.

Findings

Exposure bias is significant in SGMs for molecular conformations.

Input Perturbation effectively reduces exposure bias and enhances model performance.

State-of-the-art results are achieved on the GEOM-Drugs dataset.

Abstract

Molecular conformation generation poses a significant challenge in the field of computational chemistry. Recently, Diffusion Probabilistic Models (DPMs) and Score-Based Generative Models (SGMs) are effectively used due to their capacity for generating accurate conformations far beyond conventional physics-based approaches. However, the discrepancy between training and inference rises a critical problem known as the exposure bias. While this issue has been extensively investigated in DPMs, the existence of exposure bias in SGMs and its effective measurement remain unsolved, which hinders the use of compensation methods for SGMs, including ConfGF and Torsional Diffusion as the representatives. In this work, we first propose a method for measuring exposure bias in SGMs used for molecular conformation generation, which confirms the significant existence of exposure bias in these models and measures its value. We design a new compensation algorithm Input Perturbation (IP), which is adapted from a method originally designed for DPMs only. Experimental results show that by introducing IP, SGM-based molecular conformation models can significantly improve both the accuracy and diversity of the generated conformations. Especially by using the IP-enhanced Torsional Diffusion model, we achieve new state-of-the-art performance on the GEOM-Drugs dataset and are on par on GEOM-QM9. We provide the code publicly at https://github.com/jia-975/torsionalDiff-ip.