Synthesizable Molecular Generation via Soft-constrained GFlowNets with Rich Chemical Priors

TL;DR

This paper introduces S3-GFN, a flexible generative model that produces synthesizable molecules by soft regularization, leveraging rich chemical priors to improve drug discovery processes.

Contribution

The paper proposes S3-GFN, a novel sequence-based GFlowNet that uses soft regularization and rich priors to generate synthesizable molecules more effectively than previous hard-constrained methods.

Findings

S3-GFN achieves over 95% synthesizable molecule generation.

The model generates molecules with higher rewards across diverse tasks.

Soft regularization improves flexibility and scalability in molecular generation.

Abstract

The application of generative models for experimental drug discovery campaigns is severely limited by the difficulty of designing molecules de novo that can be synthesized in practice. Previous works have leveraged Generative Flow Networks (GFlowNets) to impose hard synthesizability constraints through the design of state and action spaces based on predefined reaction templates and building blocks. Despite the promising prospects of this approach, it currently lacks flexibility and scalability. As an alternative, we propose S3-GFN, which generates synthesizable SMILES molecules via simple soft regularization of a sequence-based GFlowNet. Our approach leverages rich molecular priors learned from large-scale SMILES corpora to steer molecular generation towards high-reward, synthesizable chemical spaces. The model induces constraints through off-policy replay training with a contrastive learning signal based on separate buffers of synthesizable and unsynthesizable samples. Our experiments show that S3-GFN learns to generate synthesizable molecules ($\geq 95\%$) with higher rewards in diverse tasks.