ProtAttn-QuadNet: An attention-based deep learning framework for protein–protein interaction prediction using ProtBERT embeddings
PLOS One · 02.06.2026 · 21 (6) · e0349433
DOI: 10.1371/journal.pone.0349433 ↗Kurz gefasst
Fachartikel in der Zeitschrift PLOS One, online erschienen am 02.06.2026. Verfasst von Md Shahidul Islam und 2 weiteren Autor:innen. Beteiligte Einrichtung: Department of Computer Science and Engineering, University of Asia Pacific, Dhaka, Bangladesh. Hauptthemen laut MEDLINE: Computational Biology, Deep Learning, Protein Interaction Mapping, Proteins. Die Arbeit nennt 29 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0).
Abstract (Originalsprache)
Protein–protein interactions (PPIs) form the backbone of most cellular processes, governing signal transduction, gene regulation, and metabolic control. However, experimental approaches to identifying PPIs remain expensive, laborious, and often incomplete. Recent advances in protein language models (PLMs) have transformed sequence-based PPI prediction by enabling deep contextual encoding of biochemical and structural information directly from amino acid sequences. Building upon this progress, we present ProtAttn-QuadNet, an attention-based deep learning framework that leverages ProtBERT embeddings to model reciprocal dependencies between protein pairs. The proposed model employs a quad-stream attention mechanism that integrates individual protein features, synergistic interactions, and complementary differences through multi-level self- and cross-attention layers. This architecture enables the discovery of fine-grained relational patterns while ensuring balanced bidirectional modeling of interacting proteins. Evaluated on the independent test set of a large-scale dataset from UniProt, ProtAttn-QuadNet achieves 97.16% accuracy (AUC-ROC 99.00%) on balanced data and 99.19% accuracy (AUC-ROC 99.76%) on oversampled datasets, surpassing several recent state-of-the-art PPI prediction methods. Statistical validation using the Chi-square and Wilcoxon signed-rank tests confirms the model’s predictive significance and reliability. ProtAttn-QuadNet offers a powerful computational framework for large-scale PPI prediction.
Schlagwörter
MeSH-Schlagwörter (MEDLINE)
Substanzen
Autor:innen und Institutionen
- Department of Computer Science and Engineering, University of Asia Pacific, Dhaka, Bangladesh.
ORCID-Profile: Md Shahidul Islam, Md Muhtasim Rahman Mim
Zitieren
APA
Islam, M. S., Mim, M. M. R., & Kabir, M. R. (2026). ProtAttn-QuadNet: An attention-based deep learning framework for protein–protein interaction prediction using ProtBERT embeddings. PLOS One, 21(6), e0349433. https://doi.org/10.1371/journal.pone.0349433
Vancouver
Islam MS, Mim MMR, Kabir MR. ProtAttn-QuadNet: An attention-based deep learning framework for protein–protein interaction prediction using ProtBERT embeddings. PLOS One. 2026;21(6):e0349433. doi:10.1371/journal.pone.0349433
BibTeX
@article{Islam2026,
author = {Islam, Md Shahidul and Mim, Md Muhtasim Rahman and Kabir, Md Raihan},
title = {ProtAttn-QuadNet: An attention-based deep learning framework for protein–protein interaction prediction using ProtBERT embeddings},
journal = {PLOS One},
year = {2026},
volume = {21},
number = {6},
pages = {e0349433},
doi = {10.1371/journal.pone.0349433}
}
EUVIMED