PhAI: A deep-learning approach to solve the crystallographic phase problem
Science · 02.08.2024 · 385 (6708) · 522-528
DOI: 10.1126/science.adn2777 ↗Kurz gefasst
Fachartikel in der Zeitschrift Science, erschienen am 02.08.2024. Verfasst von Anders S Larsen und 2 weiteren Autor:innen. Beteiligte Einrichtung: Department of Pharmacy, University of Copenhagen, Copenhagen, Denmark. Die Arbeit wurde bisher 29-mal zitiert (Crossref) und nennt 59 Literaturangaben.
Abstract (Originalsprache)
X-ray crystallography provides a distinctive view on the three-dimensional structure of crystals. To reconstruct the electron density map, the complex structure factors F = F exp i ϕ of a sufficiently large number of diffracted reflections must be known. In a conventional experiment, only the amplitudes F are obtained, and the phases ϕ are lost. This is the crystallographic phase problem. In this work, we show that a neural network, trained on millions of artificial structure data, can solve the phase problem at a resolution of only 2 angstroms, using only 10 to 20% of the data needed for direct methods. The network works in common space groups and for modest unit-cell dimensions and suggests that neural networks could be used to solve the phase problem in the general case for weakly scattering crystals.
Autor:innen und Institutionen
- Department of Pharmacy, University of Copenhagen, Copenhagen, Denmark.
ORCID-Profile: Anders S Larsen, Toms Rekis, Anders Ø Madsen
Zitieren
APA
Larsen, A. S., Rekis, T., & Madsen, A. Ø. (2024). PhAI: A deep-learning approach to solve the crystallographic phase problem. Science, 385(6708), 522-528. https://doi.org/10.1126/science.adn2777
Vancouver
Larsen AS, Rekis T, Madsen AØ. PhAI: A deep-learning approach to solve the crystallographic phase problem. Science. 2024;385(6708):522-528. doi:10.1126/science.adn2777
BibTeX
@article{Larsen2024,
author = {Larsen, Anders S and Rekis, Toms and Madsen, Anders Ø},
title = {PhAI: A deep-learning approach to solve the crystallographic phase problem},
journal = {Science},
year = {2024},
volume = {385},
number = {6708},
pages = {522--528},
doi = {10.1126/science.adn2777}
}
EUVIMED