Reconstructing continuous distributions of 3D protein structure from\n cryo-EM images

Abstract

Cryo-electron microscopy (cryo-EM) is a powerful technique for determining\nthe structure of proteins and other macromolecular complexes at near-atomic\nresolution. In single particle cryo-EM, the central problem is to reconstruct\nthe three-dimensional structure of a macromolecule from $10^{4-7}$ noisy and\nrandomly oriented two-dimensional projections. However, the imaged protein\ncomplexes may exhibit structural variability, which complicates reconstruction\nand is typically addressed using discrete clustering approaches that fail to\ncapture the full range of protein dynamics. Here, we introduce a novel method\nfor cryo-EM reconstruction that extends naturally to modeling continuous\ngenerative factors of structural heterogeneity. This method encodes structures\nin Fourier space using coordinate-based deep neural networks, and trains these\nnetworks from unlabeled 2D cryo-EM images by combining exact inference over\nimage orientation with variational inference for structural heterogeneity. We\ndemonstrate that the proposed method, termed cryoDRGN, can perform ab initio\nreconstruction of 3D protein complexes from simulated and real 2D cryo-EM image\ndata. To our knowledge, cryoDRGN is the first neural network-based approach for\ncryo-EM reconstruction and the first end-to-end method for directly\nreconstructing continuous ensembles of protein structures from cryo-EM images.\n

Cite this paper

@misc{zhong2019reconstructing,
  title = {{Reconstructing continuous distributions of 3D protein structure from\n cryo-EM images}},
  author = {Zhong, Ellen D. and Bepler, Tristan and Davis, Joseph H. and Berger, Bonnie},
  journal = {arXiv (Cornell University)},
  year = {2019},
  doi = {10.48550/arxiv.1909.05215}
}

Made from the index record. Check names, capitalisation and page numbers against the paper. Need another style? AI for Research has over 10,000.

Related papers