EUVIMEDEUVIMED entdecken
Open Access

Machine learning-driven discovery of TGFBR1 inhibitors from Mangifera indica for Atorvastatin-based antidiabetic therapy

Raghuvir Keni, Farmiza Begum, Niraja Ranadive, Pawan G Nayak, Vijayakumar Murugesan, Kommu Pradeep, Mangirish Deshpande, Nandakumar Krishnadas, Panneerselvam Theivendren iD

3 Biotech · 2026-11 · 16 (11)

DOI: 10.1007/s13205-026-05088-0 ↗

Kurz gefasst

Fachartikel in der Zeitschrift 3 Biotech, online erschienen am 12.10.2026. Verfasst von Raghuvir Keni und 8 weiteren Autor:innen. Beteiligte Einrichtung: Department of Pharmacology, Manipal College of Pharmaceutical Sciences, Manipal Academy of Higher Education, Manipal, 576104 India. Die Arbeit nennt 37 Literaturangaben. Der Volltext ist frei in PubMed Central verfügbar.

AbstractSchlagwörterBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Mangiferin, a C-glucosyl xanthone from Mangifera indica, has known antidiabetic activity, but the specific protein target underlying this activity, and its compatibility with existing antidiabetic drugs such as Atorvastatin, were not defined. In this work, machine learning, network pharmacology, molecular dynamics simulations, and experimental validation identified and validated TGFBR1 as a novel molecular target of Mangiferin and evaluated a Mangiferin-Atorvastatin combination for diabetes therapy. A curated collection of TGFBR1 inhibitors from BindingDB was used to train several supervised learning models, with the best performance from Random Forest (94.68% accuracy, κ = 0.8981) and robust external validation (79.35%) on 868 unseen compounds. Screening of 545 M. indica phytochemicals with this model identified Mangiferin, Fisetin, Isochinomin, Mangiferolic acid, and Astragalin as high-confidence TGFBR1 inhibitors. Network pharmacology identified TGF-β as the most central protein in the interaction network (cumulative score 350), and molecular docking showed that Mangiferin forms stable hydrogen bonds with SER-287, ILE-211, and GLU-208 at the TGFBR1 binding site, together with hydrophobic contacts. This binding pose remained structurally stable (RMSD 1.50-1.75 Å) over 100 nanoseconds of molecular dynamics simulation, with ligand-residue occupancy > 85%. ADMET profiling showed favorable drug-likeness (Lipinski compliance, low predicted toxicity) with a moderate permeability limitation (TPSA 201.28 Ų), and DFT analysis showed a HOMO-LUMO gap of 2.85937 eV, supporting antioxidant activity. In cultured cells, Atorvastatin combined with Mangiferin increased apoptosis through the mitochondrial pathway (Bcl-xL downregulation, BAK upregulation, cytochrome c release, caspase-7/8 activation) and enhanced insulin sensitivity, shown by increased glucose uptake (2-NBDG assay) and higher GLUT-4 and pIRS-1(Tyr896) expression. In high-fat diet-fed rats, Atorvastatin plus Mangiferin given for 24 weeks significantly improved glucose tolerance and reduced HbA1c, total cholesterol, triglycerides, and random blood sugar relative to untreated high-fat diet controls, with preserved pancreatic architecture and increased muscle GLUT-4 expression. These findings identify TGFBR1 as a plausible molecular target of Mangiferin and provide a mechanistic rationale for the Mangiferin-Atorvastatin combination as a candidate antidiabetic therapy. Supplementary Information: The online version contains supplementary material available at https://doi.org/10.1007/s13205-026-05088-0.

Schlagwörter

Stichwörter der Autor:innen

Antidiabetic therapyMachine learningMangifera indicaMolecular dynamicsNetwork pharmacologyTGFBR1 inhibitors

Autor:innen und Institutionen

  • Department of Pharmacology, Manipal College of Pharmaceutical Sciences, Manipal Academy of Higher Education, Manipal, 576104 India.
  • Department of Pharmaceutics, School of Pharmacy, JSS Academy of Higher Education and Research, Vacoas-Phoenix, Mauritius.
  • Faculty of Health Sciences, Villa College, Boduthakurufaanu Magu, Malé, Maldives.
  • Department of Pharmacology, PES's Rajaram and Tarabai Bandekar College of Pharmacy, Ponda, Goa 403401 India.
  • Department of Pharmaceutical Chemistry & Analysis, School of Pharmaceutical Sciences, Vels Institute of Science, Technology & Advanced Studies, Pallavaram, Chennai, Tamil Nadu 600117 India.

ORCID-Profile: Panneerselvam Theivendren

Zitieren

APA

Keni, R., Begum, F., Ranadive, N., Nayak, P. G., Murugesan, V., Pradeep, K., Deshpande, M., Krishnadas, N., & Theivendren, P. (2026). Machine learning-driven discovery of TGFBR1 inhibitors from Mangifera indica for Atorvastatin-based antidiabetic therapy. 3 Biotech, 16(11). https://doi.org/10.1007/s13205-026-05088-0

Vancouver

Keni R, Begum F, Ranadive N, Nayak PG, Murugesan V, Pradeep K, et al. Machine learning-driven discovery of TGFBR1 inhibitors from Mangifera indica for Atorvastatin-based antidiabetic therapy. 3 Biotech. 2026;16(11). doi:10.1007/s13205-026-05088-0

BibTeX
@article{Keni2026,
  author = {Keni, Raghuvir and Begum, Farmiza and Ranadive, Niraja and Nayak, Pawan G and Murugesan, Vijayakumar and Pradeep, Kommu and Deshpande, Mangirish and Krishnadas, Nandakumar and Theivendren, Panneerselvam},
  title = {Machine learning-driven discovery of TGFBR1 inhibitors from Mangifera indica for Atorvastatin-based antidiabetic therapy},
  journal = {3 Biotech},
  year = {2026},
  volume = {16},
  number = {11},
  doi = {10.1007/s13205-026-05088-0}
}
Kostenlos starten →