MMPCS: multi-view molecular pretraining based on consistency information and specific information
Bioinformatics · 2026
DOI: 10.1093/bioinformatics/btag028 ↗Kurz gefasst
Fachartikel in der Zeitschrift Bioinformatics, 2026. Verfasst von Chenyang Xie und 4 weiteren Autor:innen. Beteiligte Einrichtung: School of Informatics, Xiamen University, Xiamen 361005, China. Hauptthemen laut MEDLINE: Drug Discovery, Cheminformatics. Die Arbeit nennt 70 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0). Gefördert von: National Key R&D Program of China, Fujian Sunshine Charity Foundation of China.
Abstract (Originalsprache)
Motivation The goal of molecular representation learning is to automate the extraction of molecular features, a critical task in cheminformatics and drug discovery. While pretraining models using multiple views like SMILES, 2D graphs, and 3D conformations have advanced the field, integrating them effectively to produce superior representations remains a challenge. Results To bridge this gap, we propose a novel multi-view molecular pretraining method termed MMPCS, which explicitly factorizes representations into consistency and specific information. Our approach utilizes the Graph Isomorphism Network and the RoBERTa model to encode 2D molecular topological graphs and SMILES sequences, respectively. Each resulting molecular embedding is decomposed into a shared consistency component and a view-specific remainder. An autoencoder then aligns the consistency information across views. The combined consistency and view-specific representations serve as input for downstream tasks, enabling precise and task-aware predictions. When benchmarked against 16 state-of-the-art molecular pretraining methods, MMPCS achieved the highest average performance across both classification and regression tasks for molecular property prediction. It also delivered outstanding results in predicting drug-target binding affinity and cancer drug response, demonstrating its robustness and broad applicability. Additionally, a case study on the SARS-CoV-2 Omicron variant highlights the potential of MMPCS in facilitating drug repurposing efforts. Availability and implementation The source code and datasets supporting this study are publicly available at GitHub (https://github.com/xmubiocode/MMPCS) and Zenodo (https://doi.org/10.5281/zenodo.18182748).
Schlagwörter
MeSH-Schlagwörter (MEDLINE)
Autor:innen und Institutionen
- School of Informatics, Xiamen University, Xiamen 361005, China.
- Department of Advanced & Interdisciplinary Biotechnology, Academy of Military Medical Sciences, Beijing 100850, China.
ORCID-Profile: Song He, Zhongnan Zhang
Förderung
- National Key R&D Program of China
- Fujian Sunshine Charity Foundation of China
Zitieren
APA
Xie, C., Song, Y., He, S., Bo, X., & Zhang, Z. (2026). MMPCS: multi-view molecular pretraining based on consistency information and specific information. Bioinformatics. https://doi.org/10.1093/bioinformatics/btag028
Vancouver
Xie C, Song Y, He S, Bo X, Zhang Z. MMPCS: multi-view molecular pretraining based on consistency information and specific information. Bioinformatics. 2026. doi:10.1093/bioinformatics/btag028
BibTeX
@article{Xie2026,
author = {Xie, Chenyang and Song, Yingying and He, Song and Bo, Xiaochen and Zhang, Zhongnan},
title = {MMPCS: multi-view molecular pretraining based on consistency information and specific information},
journal = {Bioinformatics},
year = {2026},
doi = {10.1093/bioinformatics/btag028}
}
EUVIMED