FakET: Simulating cryo-electron tomograms with neural style transfer.

Journal: Structure (London, England : 1993)
PMID:

Abstract

In cryo-electron microscopy, accurate particle localization and classification are imperative. Recent deep learning solutions, though successful, require extensive training datasets. The protracted generation time of physics-based models, often employed to produce these datasets, limits their broad applicability. We introduce FakET, a method based on neural style transfer, capable of simulating the forward operator of any cryo transmission electron microscope. It can be used to adapt a synthetic training dataset according to reference data producing high-quality simulated micrographs or tilt-series. To assess the quality of our generated data, we used it to train a state-of-the-art localization and classification architecture and compared its performance with a counterpart trained on benchmark data. Remarkably, our technique matches the performance, boosts data generation speed 750×, uses 33× less memory, and scales well to typical transmission electron microscope detector sizes. It leverages GPU acceleration and parallel processing. The source code is available at https://github.com/paloha/faket/.

Authors

  • Pavol Harar
    Mathematical Data Science (MDS), Faculty of Mathematics, University of Vienna, Vienna, Austria; Haselbach Lab, Research Institute of Molecular Pathology (IMP), Vienna, Austria; Research Network Data Science, University of Vienna, Vienna, Austria; Department of Telecommunications, Faculty of Electrical Engineering and Communication, Brno University of Technology, Brno, Czech Republic; Institute of Science and Technology Austria (ISTA), Klosterneuburg, Austria. Electronic address: pavol.harar@ista.ac.at.
  • Lukas Herrmann
    Johann Radon Institute for Computational and Applied Mathematics, Austrian Academy of Sciences, Linz, Austria.
  • Philipp Grohs
    Johann Radon Institute, Altenberger Straße 69, A-4040 Linz, Austria; Faculty of Mathematics, University of Vienna, Oskar-Morgenstern-Platz 1, A-1090 Vienna, Austria; Research Platform Data Science @ Uni Vienna, Währinger Straße 29/S6, A-1090 Vienna, Austria. Electronic address: philipp.grohs@univie.ac.at.
  • David Haselbach
    Haselbach Lab, Research Institute of Molecular Pathology (IMP), Vienna, Austria. Electronic address: david.haselbach@imp.ac.at.