Our new X account is live! Follow @wizwand_team for updates
WorkDL logo mark

MolTC: Towards Molecular Relational Modeling In Language Models

About

Molecular Relational Learning (MRL), aiming to understand interactions between molecular pairs, plays a pivotal role in advancing biochemical research. Recently, the adoption of large language models (LLMs), known for their vast knowledge repositories and advanced logical inference capabilities, has emerged as a promising way for efficient and effective MRL. Despite their potential, these methods predominantly rely on the textual data, thus not fully harnessing the wealth of structural information inherent in molecular graphs. Moreover, the absence of a unified framework exacerbates the issue of information underutilization, as it hinders the sharing of interaction mechanism learned across diverse datasets. To address these challenges, this work proposes a novel LLM-based multi-modal framework for Molecular inTeraction prediction following Chain-of-Thought (CoT) theory, termed MolTC, which effectively integrate graphical information of two molecules in pair. To train MolTC efficiently, we introduce a Multi-hierarchical CoT concept to refine its training paradigm, and conduct a comprehensive Molecular Interactive Instructions dataset for the development of biochemical LLMs involving MRL. Our experiments, conducted across various datasets involving over 4,000,000 molecular pairs, exhibit the superiority of our method over current GNN and LLM-based baselines. Code is available at https://github.com/MangoKiller/MolTC.

Junfeng Fang, Shuai Zhang, Chang Wu, Zhengyi Yang, Zhiyuan Liu, Sihang Li, Kun Wang, Wenjie Du, Xiang Wang• 2024

Related benchmarks

TaskDatasetResultRank
Quantitative Solute-Solvent InteractionFreeSolv (test)
RMSE0.684
29
Drug-Drug Interaction predictionTWOSIDES
AUROC0.9902
13
Qualitative Drug-Drug Interaction PredictionDrugbank
Accuracy95.98
13
Qualitative Drug-Drug Interaction PredictionZhangDDI
Accuracy89.4
13
Qualitative Drug-Drug Interaction PredictionChChMiner
Accuracy95.59
13
Qualitative Drug-Drug Interaction PredictionDeepDDI
Accuracy96.7
13
Quantitative interaction predictionCombiSolv-QM (test)
MAE0.072
13
Quantitative Solute-Solvent InteractionAbraham (test)
MAE0.194
13
Quantitative Solute-Solvent InteractionCompSol (test)
MAE0.171
13
Quantitative Solute-Solvent InteractionCombiSolv (test)
MAE0.172
13
Showing 10 of 22 rows

Other info

Code

Follow for update