EUVIMEDEUVIMED entdecken
Open Access · CC BY 4.0

GAMT-GINE: A Graph Isomorphism Network Integrating Continuous Spatial Awareness and Multi-Task Learning for Protein–Ligand Binding Affinity Prediction

Jiarui Li, Hongquan Li, Di Wu, Wei He, Weinan Cao iD, Hua Yang, Zhen Hou iD

International Journal of Molecular Sciences · 12.08.2026 · 27 (16) · 7196

DOI: 10.3390/ijms27167196 ↗

Kurz gefasst

Fachartikel in der Zeitschrift International Journal of Molecular Sciences, online erschienen am 12.08.2026. Verfasst von Jiarui Li und 6 weiteren Autor:innen. Beteiligte Einrichtung: School of Information Science and Engineering, Shanxi Agricultural University, Jinzhong 030801, China. Hauptthemen laut MEDLINE: Proteins. Die Arbeit nennt 36 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0). Gefördert von: Department of Human Resources and Social Security of Shanxi Province, Shanxi Key Laboratory for Modernization of Traditional Chinese Veterinary Medicine, Shanxi Agricultural University.

AbstractSchlagwörterBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Protein–ligand interactions (PLIs) play a crucial role in drug discovery, and accurately predicting protein–ligand binding affinity (PLA) remains a central challenge in computer-aided drug design. Although graph neural networks (GNNs) have demonstrated considerable potential in molecular modeling, existing methods still face several limitations, including excessive reliance on hand-crafted chemical features, loss of spatial information, and difficulties in integrating heterogeneous affinity labels, which restrict their generalization capability in PLA prediction. To address these challenges, we propose GAMT-GINE, a graph isomorphism network that integrates continuous spatial awareness with multi-task learning. The model employs minimalist atomic features and a batch-normalization-free mechanism, together with a multi-task branch that uses a large amount of half-maximal inhibitory concentration (IC50) data as an auxiliary prediction target. Experimental results show that GAMT-GINE achieves a Pearson’s correlation coefficient (Rp) of 0.791 and a root mean square error (RMSE) of 1.403 on the CASF-2013 benchmark dataset. In the generalization evaluation on CASF-2016, Rp further increases to 0.831, while RMSE decreases to 1.227, demonstrating performance comparable to that of current State-of-the-Art models. Furthermore, comprehensive evaluations, including ablation studies, feature importance analysis, analysis of the effects of data filtering on model performance and data composition, and analysis of the influence of training–test data similarity on prediction results, indicate that GAMT-GINE can effectively utilize continuous spatial information and heterogeneous affinity labels, achieving good predictive accuracy and cross-dataset generalization capability.

Schlagwörter

MeSH-Schlagwörter (MEDLINE)

ProteinsGraph Neural NetworksLigandsProtein BindingModels, MolecularDrug Discovery

Stichwörter der Autor:innen

continuous spatial perceptiondeep learninggraph isomorphism networkmulti-task learningprotein–ligand affinity

Substanzen

LigandsProteins

Autor:innen und Institutionen

  • School of Information Science and Engineering, Shanxi Agricultural University, Jinzhong 030801, China.
  • Shanxi Key Lab for Modernization of TCVM, College of Veterinary Medicine, Shanxi Agricultural University, Jinzhong 030801, China.

ORCID-Profile: Weinan Cao, Zhen Hou

Förderung

  • Department of Human Resources and Social Security of Shanxi Province
  • Shanxi Key Laboratory for Modernization of Traditional Chinese Veterinary Medicine
  • Shanxi Agricultural University
  • Shanxi Province Doctoral Graduates and Postdoctoral Researchers Working in Shanxi Reward Research Project
  • Open Fund of the Shanxi Key Laboratory for Modernization of Traditional Chinese Veterinary Medicine

Zitieren

APA

Li, J., Li, H., Wu, D., He, W., Cao, W., Yang, H., & Hou, Z. (2026). GAMT-GINE: A Graph Isomorphism Network Integrating Continuous Spatial Awareness and Multi-Task Learning for Protein–Ligand Binding Affinity Prediction. International Journal of Molecular Sciences, 27(16), 7196. https://doi.org/10.3390/ijms27167196

Vancouver

Li J, Li H, Wu D, He W, Cao W, Yang H, et al. GAMT-GINE: A Graph Isomorphism Network Integrating Continuous Spatial Awareness and Multi-Task Learning for Protein–Ligand Binding Affinity Prediction. International Journal of Molecular Sciences. 2026;27(16):7196. doi:10.3390/ijms27167196

BibTeX
@article{Li2026,
  author = {Li, Jiarui and Li, Hongquan and Wu, Di and He, Wei and Cao, Weinan and Yang, Hua and Hou, Zhen},
  title = {GAMT-GINE: A Graph Isomorphism Network Integrating Continuous Spatial Awareness and Multi-Task Learning for Protein–Ligand Binding Affinity Prediction},
  journal = {International Journal of Molecular Sciences},
  year = {2026},
  volume = {27},
  number = {16},
  pages = {7196},
  doi = {10.3390/ijms27167196}
}
Kostenlos starten →