An image processing pipeline for electron cryo-tomography in RELION-5

Abstract

Electron tomography of frozen, hydrated samples allows structure determination of macro-molecular complexes that are embedded in complex environments. Provided that the target complexes may be localised in noisy, three-dimensional tomographic reconstructions, averaging images of multiple instances of these molecules can lead to structures with sufficient resolution for de novo atomic modelling. Although many research groups have contributed image processing tools for these tasks, a lack of standardisation and inter-operability represents a barrier for newcomers to the field. Here, we present an image processing pipeline for electron tomography data in RELION-5, with functionality ranging from the import of unprocessed movies to the automated building of atomic models in the final maps. Our explicit definition of metadata items that describe the steps of our pipeline has been designed for inter-operability with other software tools and provides a framework for further standardisation.

Nematocida displodere mechanosensitive ion channel of small conductance 2 assembles into a unique 6-channel super-structure in vitro

“…We processed the MscS2Δ2–29 cryo-EM data with Relion version 3.1 [ 53 ], 4.0 [ 54 ] and 5.0 beta [ 55 ]. Movie alignment, drift correction, and dose weighing were performed using MotionCor2 [ 56 ], and CTF estimation was done with CTFFIND (v4.1.14) [ 57 ].…”

The advent of preventive high-resolution structural histopathology by artificial-intelligence-powered cryogenic electron tomography

“…Downstream subtomogram averaging (STA) (Galaz-Montoya and Ludtke, 2017;Förster, 2022) can yield structures at subnanometer (Schur et al, 2013) and near-atomic (Schur et al, 2016) resolutions, even for complexes within cells (Tegunov et al, 2021). While STA was first demonstrated nearly 30 years ago (Walz et al, 1997) using ad hoc scripts (Nicastro et al, 2006), the first semi-automated pipelines for STA started emerging only about a decade ago (Hrabe et al, 2012;Castaño-Díez et al, 2012;Galaz-Montoya et al, 2015), and have become increasingly sophisticated and efficient (Galaz-Montoya et al, 2016;Bharat and Scheres, 2016;Castaño-Díez et al, 2017;Himes and Zhang, 2018;Fernandez and Li, 2021;Zivanov et al, 2022;Burt et al, 2024), with the field rapidly taking large strides towards full automation for some specimens (Chen et al, 2019;Ni et al, 2022;Balyschew et al, 2023). Clearly, these advancements have positioned STA at the forefront of high-resolution structural biology in situ.…”

CRISPR-Cas12a bends DNA to destabilize base pairs during target interrogation

“…In the sharpened map the DNA density was absent. Particles from the non-uniform refinement were transferred to RELION v. 5.0-beta (12) for DNA-focused refinement. After 3D refinement in RELION, signal from the protein was subtracted, the particles were re-centered on DNA, and five DNA volume classes were reconstructed.…”