From Tokens to Blocks: A Block-Diffusion Perspective on Molecular Generation

TL;DR

SoftMol introduces a novel block-diffusion framework with soft fragments and target-aware search, significantly enhancing molecular generation quality, diversity, and efficiency for drug discovery.

Contribution

It proposes SoftMol, a unified approach combining soft fragment representation, block-diffusion modeling, and target-aware search for improved molecular design.

Findings

Achieves 100% chemical validity in generated molecules.

Improves binding affinity by 9.7%.

Increases molecular diversity 2-3 times and speeds up inference 6.6 times.

Abstract

Drug discovery can be viewed as a combinatorial search over an immense chemical space, motivating the development of deep generative models for de novo molecular design. Among these, GPT-based molecular language models (MLM) have shown strong molecular design performance by learning chemical syntax and semantics from large-scale data. However, existing MLMs face two fundamental limitations: they inadequately capture the graph-structured nature of molecules when formulated as next-token prediction problems, and they typically lack explicit mechanisms for target-aware generation. Here, we propose SoftMol, a unified framework that co-designs molecular representation, model architecture, and search strategy for target-aware molecular generation. SoftMol introduces soft fragments, a rule-free block representation of SMILES that enables diffusion-native modeling, and develops SoftBD, the first block-diffusion molecular language model that combines local bidirectional diffusion with autoregressive generation under molecular structural constraints. To favor generated molecules with high drug-likeness and synthetic accessibility, SoftBD is trained on a carefully curated dataset named ZINC-Curated. SoftMol further integrates a gated Monte Carlo tree search to assemble fragments in a target-aware manner. Experimental results show that, compared with current state-of-the-art models, SoftMol achieves 100% chemical validity, improves binding affinity by 9.7%, yields a 2-3x increase in molecular diversity, and delivers a 6.6x speedup in inference efficiency. Code is available at https://github.com/szu-aicourse/softmol