EUVIMEDEUVIMED entdecken
Crossref

How good is generative diffusion model for enhanced sampling of protein conformations across scales and in all-atom resolution?

Palash Bera iD, Jagannath Mondal iD

The Journal of Chemical Physics · 21.09.2025 · 163 (11)

DOI: 10.1063/5.0279756 ↗

Kurz gefasst

Fachartikel in der Zeitschrift The Journal of Chemical Physics, online erschienen am 17.09.2025. Verfasst von Palash Bera und 1 weiteren Autor:innen. Beteiligte Einrichtung: Tata Institute of Fundamental Research Hyderabad, Hyderabad, Telangana 500046, India. Hauptthemen laut MEDLINE: Molecular Dynamics Simulation, Proteins, Intrinsically Disordered Proteins, alpha-Synuclein. Die Arbeit wurde bisher 9-mal zitiert (Crossref) und nennt 90 Literaturangaben. Gefördert von: Department of Atomic Energy, Government of India, Department of Science and Technology of India.

AbstractSchlagwörterBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Molecular dynamics (MD) simulations are fundamental for probing the structural dynamics of biomolecules, yet their efficiency is limited by the high computational cost of exploring long-timescale events. Generative machine learning (ML) models, particularly the Denoising Diffusion Probabilistic Model (DDPM), offer an emerging strategy to enhance conformational sampling. In this study, we evaluate the capabilities and limitations of DDPM in generating atomistically accurate conformational ensembles across proteins of varying size and structural order, ranging from the 20-residue folded Trp-cage and 58-residue BPTI to the 83-residue intrinsically disordered region Ash1 and the 140-residue intrinsically disordered protein α-Synuclein. Training DDPM on relatively short MD trajectories using both torsion angle and all-atom coordinate data, we demonstrate that it can reproduce key structural features such as secondary structure, radius of gyration, and contact maps, while effectively sampling sparsely populated regions of the conformational landscape. Notably, DDPM can also generate novel conformations, including transitions not explicitly observed in the training data. However, the model occasionally overlooks low-probability regions and may produce conformers with unclear physical relevance, warranting independent validation. These limitations are particularly evident in flexible systems such as IDPs. Overall, this work benchmarks DDPM as a viable tool for augmenting MD simulations, offering enhanced sampling with significant computational savings, while noting its limitations in capturing low-populated conformers. At the same time, it highlights the importance of rigorous validation and thoughtful interpretation when deploying generative models in computational biophysics.

Schlagwörter

MeSH-Schlagwörter (MEDLINE)

Molecular Dynamics SimulationProteinsIntrinsically Disordered Proteinsalpha-SynucleinProtein ConformationDiffusionMachine Learning

Substanzen

ProteinsIntrinsically Disordered Proteinsalpha-Synuclein

Autor:innen und Institutionen

  • Tata Institute of Fundamental Research Hyderabad, Hyderabad, Telangana 500046, India.

ORCID-Profile: Palash Bera, Jagannath Mondal

Förderung

  • Department of Atomic Energy, Government of India
  • Department of Science and Technology of India

Zitieren

APA

Bera, P., & Mondal, J. (2025). How good is generative diffusion model for enhanced sampling of protein conformations across scales and in all-atom resolution?. The Journal of Chemical Physics, 163(11). https://doi.org/10.1063/5.0279756

Vancouver

Bera P, Mondal J. How good is generative diffusion model for enhanced sampling of protein conformations across scales and in all-atom resolution?. The Journal of Chemical Physics. 2025;163(11). doi:10.1063/5.0279756

BibTeX
@article{Bera2025,
  author = {Bera, Palash and Mondal, Jagannath},
  title = {How good is generative diffusion model for enhanced sampling of protein conformations across scales and in all-atom resolution?},
  journal = {The Journal of Chemical Physics},
  year = {2025},
  volume = {163},
  number = {11},
  doi = {10.1063/5.0279756}
}
Kostenlos starten →