EUVIMEDEUVIMED entdecken
Open Access · CC BY 4.0

PMAVP: A Mamba-Inspired Deep Learning Framework for Antiviral Peptide Identification and Functional Activity Prediction

Peiwei Wei, Weihao Su, Qingsong Qin iD, Chuliang Wei, Yi Shi iD, Guishan Zhang iD

International Journal of Molecular Sciences · 2026

DOI: 10.3390/ijms27177764 ↗

Kurz gefasst

Fachartikel in der Zeitschrift International Journal of Molecular Sciences, 2026. Verfasst von Peiwei Wei und 5 weiteren Autor:innen. Beteiligte Einrichtung: Laboratory of Human Virology and Oncology, Shantou University Medical College, Shantou 515041, China. Hauptthemen laut MEDLINE: Peptides, Deep Learning, Antiviral Agents, Computational Biology. Die Arbeit nennt 35 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0). Gefördert von: Guangdong Basic and Applied Basic Research Foundation, Shantou Municipal Science and Technology Bureau, National Natural Science Foundation of China.

AbstractSchlagwörterBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Accurate computational prediction of antiviral peptides (AVPs) can accelerate peptide screening and reduce experimental costs. However, existing deep learning-based methods still suffer from severe class imbalance, over-reliance on handcrafted features and limited interpretability. Here, we propose PMAVP, a multi-task learning framework that integrates the ProtT5 pre-trained protein language model with a Mamba-inspired module for AVP identification and functional activity prediction. We use ProtT5 to extract deep semantic representations from peptide sequences and a Mamba module to capture long-range dependencies at a lower computational complexity. We introduce Focal Loss to mitigate class imbalance and leverage transfer learning to enhance performance on functional activity prediction. Experimental results demonstrate that our model achieves superior performance in terms of prediction accuracy, stability, and computational efficiency. Furthermore, DeepSHAP-based interpretability analysis reveals that the first 40 amino acid residues contribute substantially to AVP prediction.

Schlagwörter

MeSH-Schlagwörter (MEDLINE)

PeptidesDeep LearningAntiviral AgentsComputational BiologyPrediction AlgorithmsAmino Acid SequencePredictive Learning Models

Stichwörter der Autor:innen

DeepSHAPMambaantiviral peptidemulti-task learning

Substanzen

PeptidesAntiviral Agents

Autor:innen und Institutionen

  • Laboratory of Human Virology and Oncology, Shantou University Medical College, Shantou 515041, China.
  • Joint Shantou International Eye Center of Shantou University and the Chinese University of Hong Kong, Shantou 515041, China.
  • Department of Electronic Engineering, College of Engineering, Shantou University, Shantou 515063, China.

ORCID-Profile: Qingsong Qin, Yi Shi, Guishan Zhang

Förderung

  • Guangdong Basic and Applied Basic Research Foundation
  • Shantou Municipal Science and Technology Bureau
  • National Natural Science Foundation of China

Zitieren

APA

Wei, P., Su, W., Qin, Q., Wei, C., Shi, Y., & Zhang, G. (2026). PMAVP: A Mamba-Inspired Deep Learning Framework for Antiviral Peptide Identification and Functional Activity Prediction. International Journal of Molecular Sciences. https://doi.org/10.3390/ijms27177764

Vancouver

Wei P, Su W, Qin Q, Wei C, Shi Y, Zhang G. PMAVP: A Mamba-Inspired Deep Learning Framework for Antiviral Peptide Identification and Functional Activity Prediction. International Journal of Molecular Sciences. 2026. doi:10.3390/ijms27177764

BibTeX
@article{Wei2026,
  author = {Wei, Peiwei and Su, Weihao and Qin, Qingsong and Wei, Chuliang and Shi, Yi and Zhang, Guishan},
  title = {PMAVP: A Mamba-Inspired Deep Learning Framework for Antiviral Peptide Identification and Functional Activity Prediction},
  journal = {International Journal of Molecular Sciences},
  year = {2026},
  doi = {10.3390/ijms27177764}
}
Kostenlos starten →