VQ-SAD: Vector Quantized Structure Aware Diffusion For Molecule Generation

arXiv cs.AI / 5/4/2026

💬 OpinionIdeas & Deep AnalysisModels & Research

共有:

Key Points

The paper proposes VQ-SAD, a diffusion-based molecule generation method that incorporates symbolic molecule information rather than relying on one-hot atom and bond encodings.
VQ-SAD learns discrete latent representations for atom and bond types using a pre-trained VQ-VAE, then uses the VQ codebooks as tokenizers for the downstream diffusion process.
By treating atom and bond codes as tokens, the method aims to avoid problems like information loss in continuous embeddings and hash collisions from fingerprint-based approaches.
VQ-SAD is presented as a neuro-symbolic model with a learnable forward diffusion process and a larger discrete code space to improve the denoising balance across atom and bond types.
Experimental results on QM9 and ZINC250k show that VQ-VAE (within the proposed approach) slightly outperforms state-of-the-art diffusion-based molecule generation methods.

Abstract

Many diffusion based molecule generation methods ignore the symbolic information of molecules and represent the atom and bond type as one hot representation. Methods based on Morgan fingerprints produce hash collisions and are hard to embed into a continuous space without information loss and random fingerprints correspond to no valid molecule. To circumvent this issue we use another paradigm and consider atom and bond codes as latent variables of VQ-VAE. We introduce VQ-SAD which first trains a VQ-VAE and uses the frozen pretrained VQ-VAE model and considers the codebooks for both atom and bond types as tokenizers for the downstream diffusion process. VQ-SAD is a neuro-symbolic model that utilizes both symbolic and neural structural information for a diffusion based model with learnable forward process. The large discrete code space provides a more balanced atom and bond types which enhances the denoising process. VQ-VAE slightly outperforms SOTA models for diffusion based molecule generation on QM9 and ZINC250k datasets.