EUVIMEDEUVIMED entdecken
Open Access · CC BY 4.0

Deep residual networks for crystallography trained on synthetic data

Derek Mendez, James M Holton, Artem Y Lyubimov, Sabine Hollatz, Irimpan I Mathews iD, Aleksander Cichosz, Vardan Martirosyan, Teo Zeng, Ryan Stofer, Ruobin Liu iD, Jinhu Song, Scott McPhillips und 2 weitere

Acta Crystallographica Section D Structural Biology · 2024

DOI: 10.1107/s2059798323010586 ↗

Kurz gefasst

Fachartikel in der Zeitschrift Acta Crystallographica Section D Structural Biology, 2024. Verfasst von Derek Mendez und 13 weiteren Autor:innen. Beteiligte Einrichtung: Stanford Synchrotron Radiation Lightsource, SLAC National Accelerator Laboratory, Menlo Park, CA 94025, USA. Hauptthemen laut MEDLINE: Artificial Intelligence, Synchrotrons. Die Arbeit wurde bisher 3-mal zitiert (Crossref) und nennt 64 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0). Gefördert von: NIGMS NIH HHS, National Institutes of Health, National Institute of General Medical Sciences, National Science Foundation.

AbstractSchlagwörterBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

The use of artificial intelligence to process diffraction images is challenged by the need to assemble large and precisely designed training data sets. To address this, a codebase called Resonet was developed for synthesizing diffraction data and training residual neural networks on these data. Here, two per-pattern capabilities of Resonet are demonstrated: (i) interpretation of crystal resolution and (ii) identification of overlapping lattices. Resonet was tested across a compilation of diffraction images from synchrotron experiments and X-ray free-electron laser experiments. Crucially, these models readily execute on graphics processing units and can thus significantly outperform conventional algorithms. While Resonet is currently utilized to provide real-time feedback for macromolecular crystallography users at the Stanford Synchrotron Radiation Lightsource, its simple Python-based interface makes it easy to embed in other processing frameworks. This work highlights the utility of physics-based simulation for training deep neural networks and lays the groundwork for the development of additional models to enhance diffraction collection and analysis.

Schlagwörter

MeSH-Schlagwörter (MEDLINE)

Artificial IntelligenceSynchrotronsCrystallography, X-RayAlgorithmsComputer Simulation

Stichwörter der Autor:innen

XFELsartificial intelligencerotation crystallographyserial crystallographysynchrotrons

Autor:innen und Institutionen

  • Stanford Synchrotron Radiation Lightsource, SLAC National Accelerator Laboratory, Menlo Park, CA 94025, USA.
  • Department of Statistics and Applied Probability, UC Santa Barbara, Santa Barbara, CA 93106, USA.
  • Department of Mathematics, UC Santa Barbara, Santa Barbara, CA 93106, USA.

ORCID-Profile: Irimpan I Mathews, Ruobin Liu, Aina E Cohen

Förderung

  • NIGMS NIH HHS
  • National Institutes of Health, National Institute of General Medical Sciences
  • National Science Foundation
  • SLAC National Accelerator Laboratory

Zitieren

APA

Mendez, D., Holton, J. M., Lyubimov, A. Y., Hollatz, S., Mathews, I. I., Cichosz, A., Martirosyan, V., Zeng, T., Stofer, R., Liu, R., Song, J., McPhillips, S., Soltis, M., & Cohen, A. E. (2024). Deep residual networks for crystallography trained on synthetic data. Acta Crystallographica Section D Structural Biology. https://doi.org/10.1107/s2059798323010586

Vancouver

Mendez D, Holton JM, Lyubimov AY, Hollatz S, Mathews II, Cichosz A, et al. Deep residual networks for crystallography trained on synthetic data. Acta Crystallographica Section D Structural Biology. 2024. doi:10.1107/s2059798323010586

BibTeX
@article{Mendez2024,
  author = {Mendez, Derek and Holton, James M and Lyubimov, Artem Y and Hollatz, Sabine and Mathews, Irimpan I and Cichosz, Aleksander and Martirosyan, Vardan and Zeng, Teo and Stofer, Ryan and Liu, Ruobin and Song, Jinhu and McPhillips, Scott and Soltis, Mike and Cohen, Aina E},
  title = {Deep residual networks for crystallography trained on synthetic data},
  journal = {Acta Crystallographica Section D Structural Biology},
  year = {2024},
  doi = {10.1107/s2059798323010586}
}
Kostenlos starten →