EUVIMEDEUVIMED entdecken
Open Access · CC BY 4.0

GRAPE: graph-regularized protein language modeling unlocks TCR-epitope binding specificity

Xiangzheng Fu iD, Li Peng iD, Haowen Chen iD, Mingqiang Rong, Yifan Chen, Dongsheng Cao iD, Sisi Yuan, Aiping Lu

Briefings in Bioinformatics · 2025

DOI: 10.1093/bib/bbaf522 ↗

Kurz gefasst

Fachartikel in der Zeitschrift Briefings in Bioinformatics, 2025. Verfasst von Xiangzheng Fu und 7 weiteren Autor:innen. Beteiligte Einrichtung: Institute of Artificial Intelligence Application, College of Computer and Information Engineering, Central South University of Forestry and Technolog…. Hauptthemen laut MEDLINE: Epitopes, T-Lymphocyte, Receptors, Antigen, T-Cell. Die Arbeit wurde bisher 18-mal zitiert (Crossref) und nennt 33 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0). Gefördert von: National Natural Science Foundation of China, Natural Science Foundation of Hunan Province, Educational Commission of Hunan Province.

AbstractSchlagwörterBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

T-cell receptor (TCR)-epitope binding prediction is critical for immunotherapies but remains challenged by sparse interaction networks and severe class imbalance in training data. Current graph neural network (GNN) approaches for predicting TCR-epitope binding (TEB) fail to address two key limitations: over-smoothing during message propagation in sparse TCR-epitope graphs and biased predictions toward dominant epitope-TCR pairs. Here, we present GRAPE (Graph-Regularized Attentive Protein Embeddings), a framework unifying spectral graph regularization and imbalance-aware learning. GRAPE first leverages protein language models (ESM-2) to generate evolutionary-informed TCR/epitope embeddings, constructing a topology-aware interaction graph. To mitigate over-smoothing, we introduce spectral graph regularization, explicitly constraining node feature smoothness to preserve discriminative patterns in sparse neighborhoods. Simultaneously, a dynamic edge reweighting module prioritizes unobserved TCR-epitope edges during graph propagation, coupled with a differentiable area under the ROC curve-maximization objective that directly optimizes for imbalance resilience. Extensive benchmarking on public datasets demonstrates that GRAPE significantly outperforms state-of-the-art methods in TEB prediction. This work establishes GRAPE as a robust framework for elucidating TCR-epitope interactions, with broad applications in immunology research and therapeutic design.

Schlagwörter

MeSH-Schlagwörter (MEDLINE)

Epitopes, T-LymphocyteReceptors, Antigen, T-CellProtein BindingSoftwareComputational BiologyHumansAnimalsDatasets as TopicAmino Acid Sequence

Stichwörter der Autor:innen

AUC-maximizationTCR-epitope bindinggraph regularizationprotein language models

Substanzen

Epitopes, T-LymphocyteReceptors, Antigen, T-Cell

Autor:innen und Institutionen

  • Institute of Artificial Intelligence Application, College of Computer and Information Engineering, Central South University of Forestry and Technology, No. 498 Shaoshan South Road, Tianxin District, Changsha, Hunan 4100…
  • School of Chinese Medicine, Hong Kong Baptist University, 15 Baptist University Road, Kowloon Tong, Kowloon, Hong Kong SAR 999077, China.
  • College of Computer Science and Engineering, Hunan University of Science and Technology, No. 1 Taoyuan Road, Yuhu District, Xiangtan, Hunan 411201, China.
  • College of Science and Electronic Engineering, Hunan University, 2 Lushan South Road, Yuelu District, Changsha, Hunan 410082, China.
  • The National & Local Joint Engineering Laboratory of Animal Peptide Drug Development, College of Life Sciences, Hunan Normal University, 36 Lushan Road, Yuelu District, Changsha, Hunan 410081, China.
  • School of Information Engineering, Changsha Medical University, No. 1501 Leifeng Road, Wangcheng District, Changsha, 410219, China.

ORCID-Profile: Xiangzheng Fu, Li Peng, Haowen Chen, Dongsheng Cao

Förderung

  • National Natural Science Foundation of China
  • Natural Science Foundation of Hunan Province
  • Educational Commission of Hunan Province

Zitieren

APA

Fu, X., Peng, L., Chen, H., Rong, M., Chen, Y., Cao, D., Yuan, S., & Lu, A. (2025). GRAPE: graph-regularized protein language modeling unlocks TCR-epitope binding specificity. Briefings in Bioinformatics. https://doi.org/10.1093/bib/bbaf522

Vancouver

Fu X, Peng L, Chen H, Rong M, Chen Y, Cao D, et al. GRAPE: graph-regularized protein language modeling unlocks TCR-epitope binding specificity. Briefings in Bioinformatics. 2025. doi:10.1093/bib/bbaf522

BibTeX
@article{Fu2025,
  author = {Fu, Xiangzheng and Peng, Li and Chen, Haowen and Rong, Mingqiang and Chen, Yifan and Cao, Dongsheng and Yuan, Sisi and Lu, Aiping},
  title = {GRAPE: graph-regularized protein language modeling unlocks TCR-epitope binding specificity},
  journal = {Briefings in Bioinformatics},
  year = {2025},
  doi = {10.1093/bib/bbaf522}
}
Kostenlos starten →