GraphEBM: Molecular Graph Generation with Energy-Based Models
This addresses the issue of bias in generative models for molecular design, which is important for drug discovery and materials science, but it appears incremental as it builds on existing energy-based models.
The paper tackles the problem of molecular graph generation by proposing GraphEBM, an energy-based model that ensures permutation invariance to avoid bias, and it demonstrates effectiveness in random, goal-directed, and compositional generation tasks.
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.