GraphEBM: Molecular Graph Generation with Energy-Based Models

TL;DR

GraphEBM introduces a permutation-invariant energy-based model for molecular graph generation, enabling property-driven and multi-objective molecule synthesis with promising experimental results.

Contribution

The paper presents a novel permutation-invariant energy-based model for molecular graph generation, incorporating property-guided sampling and multi-objective capabilities.

Findings

Effective generation of molecules with desired properties.

Permutation invariance improves generation bias.

Versatile multi-objective molecule synthesis.

Abstract

We note that most existing approaches for molecular graph generation fail to guarantee the intrinsic property of permutation invariance, resulting in unexpected bias in generative models. In this work, we propose GraphEBM to generate molecular graphs using energy-based models. In particular, we parameterize the energy function in a permutation invariant manner, thus making GraphEBM permutation invariant. We apply Langevin dynamics to train the energy function by approximately maximizing likelihood and generate samples with low energies. Furthermore, to generate molecules with a desirable property, we propose a simple yet effective strategy, which pushes down energies with flexible degrees according to the properties of corresponding molecules. Finally, we explore the use of GraphEBM for generating molecules with multiple objectives in a compositional manner. Comprehensive experimental results on random, goal-directed, and compositional generation tasks demonstrate the effectiveness of our proposed method.