EUVIMEDEUVIMED entdecken
Open Access · CC BY 4.0

Multi-Modal Topology-Aware Graph Neural Network for Robust Chemical–Protein Interaction Prediction

Jianshi Wang iD

International Journal of Molecular Sciences · 2025

DOI: 10.3390/ijms26178666 ↗

Kurz gefasst

Fachartikel in der Zeitschrift International Journal of Molecular Sciences, 2025. Verfasst von Jianshi Wang. Beteiligte Einrichtung: Department of Systems Innovation, Graduate School of Engineering, Hongo Campus, The University of Tokyo, Tokyo 113-8656, Japan. Hauptthemen laut MEDLINE: Neural Networks, Computer, Proteins, Protein Interaction Mapping, Protein Interaction Maps. Die Arbeit wurde bisher 10-mal zitiert (Crossref) und nennt 38 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0). Gefördert von: Os'Lab Co.,ltd, Oda Pharmaceuticals.Corp, Japan Science and Technology Agency.

AbstractSchlagwörterBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Reliable prediction of chemical–protein interactions (CPIs) remains a key challenge in drug discovery, especially under sparse or noisy biological data. We present MM-TCoCPIn, a Multi-Modal Topology-aware Chemical–Protein Interaction Network that integrates three causally grounded modalities—network topology, biomedical semantics, and a 3D protein structure—into an interpretable graph learning framework. The model processes topological features via a CTC (Comprehensive Topological Characteristics)-based encoder, literature-derived semantics via SciBERT (Scientific Bidirectional Encoder Representations from Transformers), and structural geometry via a GVP-GNN (Geometric Vector Perceptron Graph Neural Network) applied to AlphaFold2 contact graphs. Evaluation on datasets from STITCH, STRING, and PubMed shows that MM-TCoCPIn achieves state-of-the-art performance (AUC = 0.93, F1 = 0.92), outperforming uni-modal baselines. Importantly, ablation and counterfactual analyses confirm that each modality contributes distinct biological insight: topology ensures robustness, semantics enhance recall, and structure sharpens precision. This framework offers a scalable and causally interpretable solution for CPI modeling, bridging the gap between predictive accuracy and mechanistic understanding.

Schlagwörter

MeSH-Schlagwörter (MEDLINE)

Neural Networks, ComputerProteinsProtein Interaction MappingProtein Interaction MapsHumansAlgorithmsDrug DiscoveryGraph Neural Networks

Stichwörter der Autor:innen

causal interpretabilitychemical–protein interaction predictionmulti-modal graph neural networktopological reasoning

Substanzen

Proteins

Autor:innen und Institutionen

  • Department of Systems Innovation, Graduate School of Engineering, Hongo Campus, The University of Tokyo, Tokyo 113-8656, Japan.
  • Os' Lab, Twin Towers South 17th Floor, 1-13-1 Umeda, Kita-ku, Osaka 530-0001, Japan.

ORCID-Profile: Jianshi Wang

Förderung

  • Os'Lab Co.,ltd
  • Oda Pharmaceuticals.Corp
  • Japan Science and Technology Agency
  • Japan Society for the Promotion of Science
  • Oda Pharmaceuticals.Corp, Os’Lab Co., ltd.
  • Q-Leap
  • JSPS Kakenhi
  • MEXT Initiative for Life Design Innovation

Zitieren

APA

Wang, J. (2025). Multi-Modal Topology-Aware Graph Neural Network for Robust Chemical–Protein Interaction Prediction. International Journal of Molecular Sciences. https://doi.org/10.3390/ijms26178666

Vancouver

Wang J. Multi-Modal Topology-Aware Graph Neural Network for Robust Chemical–Protein Interaction Prediction. International Journal of Molecular Sciences. 2025. doi:10.3390/ijms26178666

BibTeX
@article{Wang2025,
  author = {Wang, Jianshi},
  title = {Multi-Modal Topology-Aware Graph Neural Network for Robust Chemical–Protein Interaction Prediction},
  journal = {International Journal of Molecular Sciences},
  year = {2025},
  doi = {10.3390/ijms26178666}
}
Kostenlos starten →