EUVIMEDEUVIMED entdecken
Open Access

Toward ultimate NMR resolution with deep learning

Amir Jahangiri iD, Tatiana Agback iD, Ulrika Brath iD, Vladislav Orekhov iD

Science Advances · 27.03.2026 · 12 (13)

DOI: 10.1126/sciadv.ady7995 ↗

Kurz gefasst

Fachartikel in der Zeitschrift Science Advances, erschienen am 27.03.2026. Verfasst von Amir Jahangiri und 3 weiteren Autor:innen. Beteiligte Einrichtung: Department of Chemistry and Molecular Biology, University of Gothenburg, Gothenburg 40530, Sweden. Die Arbeit wurde bisher 2-mal zitiert (Crossref) und nennt 77 Literaturangaben. Der Volltext ist frei in PubMed Central verfügbar. Gefördert von: Vetenskapsrådet.

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Resolution in NMR is defined as the ability to distinguish and accurately determine signal positions while mitigating overlap. In the pursuit of ultimate resolution, we introduce peak probability presentations ( P 3 ), a statistical spectral representation that assigns a probability to each spectral point, indicating the likelihood that a peak maximum occurs at that location. The mapping between the traditional spectrum and P 3 is achieved using MR-Ai, a physics-inspired and computationally efficient deep-learning neural network. P 3 is validated on 60 database proteins and showcased on the challenging Tau and MATL1 proteins. Using synthetic spectra, we show that the achieved peak-localization precision closely approaches the theoretical limits set by the Cramér-Rao lower bound and Bayesian Monte Carlo estimates. Furthermore, MR-Ai enables the coprocessing of multiple spectra, facilitating direct information exchange between datasets to enhance spectral quality, particularly in cases of highly sparse sampling.

Autor:innen und Institutionen

  • Department of Chemistry and Molecular Biology, University of Gothenburg, Gothenburg 40530, Sweden.
  • Department of Molecular Sciences, Swedish University of Agricultural Sciences, Uppsala 75007, Sweden.
  • Swedish NMR Centre, SciLifeLab, University of Gothenburg, P.O. Box 465, Gothenburg 40530, Sweden.

ORCID-Profile: Amir Jahangiri, Tatiana Agback, Ulrika Brath, Vladislav Orekhov

Förderung

  • Vetenskapsrådet

Zitieren

APA

Jahangiri, A., Agback, T., Brath, U., & Orekhov, V. (2026). Toward ultimate NMR resolution with deep learning. Science Advances, 12(13). https://doi.org/10.1126/sciadv.ady7995

Vancouver

Jahangiri A, Agback T, Brath U, Orekhov V. Toward ultimate NMR resolution with deep learning. Science Advances. 2026;12(13). doi:10.1126/sciadv.ady7995

BibTeX
@article{Jahangiri2026,
  author = {Jahangiri, Amir and Agback, Tatiana and Brath, Ulrika and Orekhov, Vladislav},
  title = {Toward ultimate NMR resolution with deep learning},
  journal = {Science Advances},
  year = {2026},
  volume = {12},
  number = {13},
  doi = {10.1126/sciadv.ady7995}
}
Kostenlos starten →