用于逆向分子设计的多模态大型语言模型与回溯合成规划

Multimodal Large Language Models for Inverse Molecular Design with Retrosynthetic Planning

October 5, 2024

作者: Gang Liu, Michael Sun, Wojciech Matusik, Meng Jiang, Jie Chen

cs.AI

摘要

尽管大型语言模型（LLMs）已经整合了图像，但将其适应图形仍然具有挑战性，这限制了它们在材料和药物设计中的应用。这一困难源于需要在文本和图形之间实现连贯的自回归生成。为了解决这个问题，我们引入了Llamole，这是第一个能够交替生成文本和图形的多模态LLM，实现了具有逆合成规划的分子设计。Llamole将基础LLM与图扩散变压器和图神经网络相结合，用于在文本中进行多条件分子生成和反应推断，同时增强了对分子的理解，灵活地控制不同图模块之间的激活。此外，Llamole将A*搜索与基于LLM的成本函数相结合，实现了高效的逆合成规划。我们创建了基准数据集，并进行了大量实验，评估了Llamole与上下文学习和监督微调的性能。在可控分子设计和逆合成规划的12个指标中，Llamole在14个改进的LLM模型中显著优于其他模型。

English

While large language models (LLMs) have integrated images, adapting them to graphs remains challenging, limiting their applications in materials and drug design. This difficulty stems from the need for coherent autoregressive generation across texts and graphs. To address this, we introduce Llamole, the first multimodal LLM capable of interleaved text and graph generation, enabling molecular inverse design with retrosynthetic planning. Llamole integrates a base LLM with the Graph Diffusion Transformer and Graph Neural Networks for multi-conditional molecular generation and reaction inference within texts, while the LLM, with enhanced molecular understanding, flexibly controls activation among the different graph modules. Additionally, Llamole integrates A* search with LLM-based cost functions for efficient retrosynthetic planning. We create benchmarking datasets and conduct extensive experiments to evaluate Llamole against in-context learning and supervised fine-tuning. Llamole significantly outperforms 14 adapted LLMs across 12 metrics for controllable molecular design and retrosynthetic planning.