Text-Augmented Multimodal LLMs for Chemical Reaction Condition Recommendation

ICLR 2025 Conference Submission373 Authors

Keywords: Text-augmented, Multimodal LLM, Chemical reaction condition recommendation

TL;DR: Text-Augmented Multimodal LLMs for Chemical Reaction Condition Recommendation

Abstract: High-throughput reaction condition (RC) screening is fundamental to chemical synthesis. However, current RC screening suffers from laborious and costly trial-and-error workflows. Traditional computer-aided synthesis planning (CASP) tools fail to find suitable RCs due to data sparsity and inadequate reaction representations. Nowadays, large language models (LLMs) are capable of tackling chemistry-related problems, such as molecule design, and chemical logic Q\&A tasks. However, LLMs have not yet achieved accurate predictions of chemical reaction conditions. Here, we present Chemma-RC, a text-augmented multimodal LLM that responds to task-specific questions by generating answers about reaction conditions. It learns a unified reaction representation via modality alignment from a corpus of reactions and question prompts, molecular structures in SMILES format, and graphical representations of chemical reactions. We construct a 1.2 million pair-wised Q\&A instruction dataset to train Chemma-RC and design a projection module for modality alignment. Our experimental results demonstrate that Chemma-RC achieves state-of-the-art performance on two open benchmark datasets and exhibits strong generalization capabilities on out-of-domain (OOD) and High-Throughput Experimentation (HTE) datasets. Chemma-RC has the potential to accelerate high-throughput condition screening in chemical synthesis.

Primary Area: applications to physical sciences (physics, chemistry, biology, etc.)

Code Of Ethics: I acknowledge that I and all co-authors of this work have read and commit to adhering to the ICLR Code of Ethics.

Submission Guidelines: I certify that this submission complies with the submission instructions as described on https://iclr.cc/Conferences/2025/AuthorGuide.

Reciprocal Reviewing: I understand the reciprocal reviewing requirement as described on https://iclr.cc/Conferences/2025/CallForPapers. If none of the authors are registered as a reviewer, it may result in a desk rejection at the discretion of the program chairs. To request an exception, please complete this form at https://forms.gle/Huojr6VjkFxiQsUp6.

Anonymous Url: I certify that there is no URL (e.g., github page) that could be used to find authors’ identity.

No Acknowledgement Section: I certify that there is no acknowledgement section in this submission for double blind review.

Submission Number: 373