EUVIMEDEUVIMED entdecken
Crossref

Computationally Efficient Network Models Successfully Predict Allosteric Sites of SARS ‐ CoV ‐2 Main Protease and Reveal Its Dynamic Allostery

Merve Yuce iD, Ozge Kurkcuoglu iD

Proteins: Structure, Function, and Bioinformatics · 2026

DOI: 10.1002/prot.70122 ↗

Kurz gefasst

Fachartikel in der Zeitschrift Proteins: Structure, Function, and Bioinformatics, 2026. Verfasst von Merve Yuce und 1 weiteren Autor:innen. Beteiligte Einrichtung: Department of Chemical Engineering, Istanbul Technical University, Istanbul, Turkey. Hauptthemen laut MEDLINE: SARS-CoV-2, Viral Nonstructural Proteins, Coronavirus 3C Proteases, Cysteine Endopeptidases. Die Arbeit nennt 96 Literaturangaben. Gefördert von: Scientific Research Projects Coordination Unit, Istanbul Technical University, TUBITAK National PhD Scholarship Program in the Priority Fields in Science and Technology, Bilimsel Araştırma Projeleri Birimi, İstanbul Teknik Üniversitesi.

AbstractSchlagwörterBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Developing allosteric drugs to treat pathogenic diseases can offer a promising alternative to orthosteric drugs that may bind to conserved motifs in human homologs. The allosteric drugs bind to allosteric sites, induce changes in the target protein's active site, and modulate its function with high selectivity, reduced adverse effects, and low toxicity. While identifying allosteric sites is costly and labor‐intensive with experimental approaches, computational methods utilizing three‐dimensional protein structures offer a cost‐effective solution for discovering potential allosteric sites and predicting the effects of ligand binding. This study evaluates the effectiveness of two network models, the residue interaction network (RIN) model and the mixed coarse‐grained anisotropic network model (mcgANM), in identifying putative allosteric regions, predicting the structural response of the protein to ligand binding, and elucidating allosteric mechanisms while maintaining computational efficiency. The SARS‐CoV‐2 main protease (M pro ) is employed as an allosteric protein model due to a rich experimental and computational data available since the COVID‐19 pandemic. The findings of the methods are assessed with statistical analysis, all‐atom molecular dynamics simulations, and other elastic network models, namely Essential Site Scanning Analysis and Gaussian Network Model using a dataset of 15 ligand‐bound and 4 ligand‐free structures. RIN predicted the known drug binding sites of M pro with high statistics, up to 80.0% sensitivity, 89.7% specificity, 29.6% precision, and 89.2% accuracy. RIN suggested an allosteric mechanism of M pro that facilitates the allosteric communication of the allosteric and active sites through residue fluctuations. RIN was able to decompose the enzyme structure to dynamic domains, showing the organization of structural components to form a functional viral protease. mcgANM suggested the changes in residue fluctuations after ligand binding. The findings underscore the utility of the network models in advancing allosteric drug design.

Schlagwörter

MeSH-Schlagwörter (MEDLINE)

SARS-CoV-2Viral Nonstructural ProteinsCoronavirus 3C ProteasesCysteine EndopeptidasesAllosteric SiteHumansAllosteric RegulationMolecular Dynamics SimulationLigandsProtein BindingCOVID-19Catalytic DomainComputational BiologyProtein Conformation

Stichwörter der Autor:innen

allosteryelastic network modelmixed‐coarse‐grained ANMmolecular dynamic simulationsprotein dynamicsresidue interaction network

Substanzen

LigandsViral Nonstructural ProteinsCoronavirus 3C ProteasesCysteine Endopeptidases

Autor:innen und Institutionen

  • Department of Chemical Engineering, Istanbul Technical University, Istanbul, Turkey.

ORCID-Profile: Merve Yuce, Ozge Kurkcuoglu

Förderung

  • Scientific Research Projects Coordination Unit, Istanbul Technical University
  • TUBITAK National PhD Scholarship Program in the Priority Fields in Science and Technology
  • Bilimsel Araştırma Projeleri Birimi, İstanbul Teknik Üniversitesi

Zitieren

APA

Yuce, M., & Kurkcuoglu, O. (2026). Computationally Efficient Network Models Successfully Predict Allosteric Sites of SARS ‐ CoV ‐2 Main Protease and Reveal Its Dynamic Allostery. Proteins: Structure, Function, and Bioinformatics. https://doi.org/10.1002/prot.70122

Vancouver

Yuce M, Kurkcuoglu O. Computationally Efficient Network Models Successfully Predict Allosteric Sites of SARS ‐ CoV ‐2 Main Protease and Reveal Its Dynamic Allostery. Proteins: Structure, Function, and Bioinformatics. 2026. doi:10.1002/prot.70122

BibTeX
@article{Yuce2026,
  author = {Yuce, Merve and Kurkcuoglu, Ozge},
  title = {Computationally Efficient Network Models Successfully Predict Allosteric Sites of SARS ‐ CoV ‐2 Main Protease and Reveal Its Dynamic Allostery},
  journal = {Proteins: Structure, Function, and Bioinformatics},
  year = {2026},
  doi = {10.1002/prot.70122}
}
Kostenlos starten →