Abstract:Cryo-electron tomography (cryo-ET) combined with sub-tomogram averaging (STA) allows the determination of protein structures imaged within the native context of the cell at near-atomic resolution. Particle picking is an essential step in the cryo-ET/STA image analysis pipeline that consists in locating the position of proteins within crowded cellular tomograms so that they can be aligned and averaged in 3D to improve resolution. While extensive work in 2D particle picking has been done in the context of single-particle cryo-EM, comparatively fewer strategies have been proposed to pick particles from 3D tomograms, in part due to the challenges associated with working with noisy 3D volumes affected by the missing wedge. While strategies based on 3D template-matching and deep learning are commonly used, these methods are computationally expensive and require either an external template or manual labelling which can bias the results and limit their applicability. Here, we propose a size-based method to pick particles from tomograms that is fast, accurate, and does not require external templates or user provided labels. We compare the performance of our approach against a commonly used algorithm based on deep learning, crYOLO, and show that our method: i) has higher detection accuracy, ii) does not require user input for labeling or time-consuming training, and iii) runs efficiently on non-specialized CPU hardware. We demonstrate the effectiveness of our approach by automatically detecting particles from tomograms representing different types of samples and using these particles to determine the high-resolution structures of ribosomes imaged in vitro and in situ.

Adaptive Tilt-Series Alignment with Feature Resampling in Cryo-Electron Tomography

Tilt-series-based joint CTF estimation for cryo-electron tomography

Markerauto2: A fast and robust fully automatic fiducial marker-based tilt series alignment software for electron tomography

Fine-grained alignment of cryo-electron subtomograms based on MPI parallel optimization

An approach for coherent periodogram averaging of tilt-series data for improved CTF estimation

TomoAlign: A novel approach to correcting sample motion and 3D CTF in CryoET

Automated fiducial-based alignment of cryo-electron tomography tilt series in Dynamo

SparseAlign: A Super-Resolution Algorithm for Automatic Marker Localization and Deformation Estimation in Cryo-Electron Tomography

Structure Detection in Three-Dimensional Cellular Cryoelectron Tomograms by Reconstructing Two-Dimensional Annotated Tilt Series

A deep learning approach to the automatic detection of alignment errors in cryo-electron tomographic reconstructions

Implementation of a cryo-electron tomography tilt-scheme optimized for high resolution subtomogram averaging

A Simulated Annealing Approach for Resolution Guided Homogeneous Cryo-Electron Microscopy Image Selection.

Automatic Localization and Identification of Mitochondria in Cellular Electron Cryo-Tomography Using Faster-Rcnn

Protein complex structure modeling by cross-modal alignment between cryo-EM maps and protein sequences

What shapes template matching performance in cryogenic electron tomography

A Complete Data Processing Workflow for Cryo-Et and Subtomogram Averaging

Accurate size-based protein localization from cryo-ET tomograms

goCTF: Geometrically optimized CTF determination for single-particle cryo-EM

Fast Cryo-EM Image Alignment Algorithm Using Power Spectrum Features

Computational Methods Toward Unbiased Pattern Mining and Structure Determination in Cryo-Electron Tomography Data

Computational Methods for in Situ Structural Studies with Cryogenic Electron Tomography.