Can LLMs Generate Diverse Molecules? Towards Alignment with Structural Diversity

TL;DR

This paper introduces a novel fine-tuning and reinforcement learning approach for large language models to generate structurally diverse molecules, addressing the limitation of current models that tend to produce similar molecules, thereby enhancing drug discovery potential.

Contribution

The paper presents a new method combining supervised fine-tuning and reinforcement learning to improve the structural diversity of molecules generated by LLMs.

Findings

Enhanced molecular diversity compared to existing methods

Effective autoregressive generation of diverse molecules

Improved alignment of textual and structural diversity

Abstract

Recent advancements in large language models (LLMs) have demonstrated impressive performance in molecular generation, which offers potential to accelerate drug discovery. However, the current LLMs overlook a critical requirement for drug discovery: proposing a diverse set of molecules. This diversity is essential for improving the chances of finding a viable drug, as it provides alternative molecules that may succeed where others fail in real-world validations. Nevertheless, the LLMs often output structurally similar molecules. While decoding schemes like diverse beam search may enhance textual diversity, this often does not align with molecular structural diversity. In response, we propose a new method for fine-tuning molecular generative LLMs to autoregressively generate a set of structurally diverse molecules, where each molecule is generated by conditioning on the previously generated molecules. Our approach consists of two stages: (1) supervised fine-tuning to adapt LLMs to autoregressively generate molecules in a sequence and (2) reinforcement learning to maximize structural diversity within the generated molecules. Our experiments show that the proposed approach enables LLMs to generate diverse molecules better than existing approaches for diverse sequence generation.