Crossing New Frontiers: Knowledge-Augmented Large Language Model Prompting for Zero-Shot Text-Based De Novo Molecule Design

TL;DR

This paper introduces a knowledge-augmented prompting method for large language models to enable zero-shot, text-based de novo molecule design, significantly improving generation quality over existing models.

Contribution

It presents a novel prompting framework that incorporates task-specific instructions and demonstrations, effectively addressing distributional shift in molecule generation tasks.

Findings

Outperforms state-of-the-art baseline models on benchmark datasets.

Effective in zero-shot molecular generation with technical descriptions.

Addresses distributional shift challenges in LLM-based molecule design.

Abstract

Molecule design is a multifaceted approach that leverages computational methods and experiments to optimize molecular properties, fast-tracking new drug discoveries, innovative material development, and more efficient chemical processes. Recently, text-based molecule design has emerged, inspired by next-generation AI tasks analogous to foundational vision-language models. Our study explores the use of knowledge-augmented prompting of large language models (LLMs) for the zero-shot text-conditional de novo molecular generation task. Our approach uses task-specific instructions and a few demonstrations to address distributional shift challenges when constructing augmented prompts for querying LLMs to generate molecules consistent with technical descriptions. Our framework proves effective, outperforming state-of-the-art (SOTA) baseline models on benchmark datasets.