Advances in Software Methods and Tools for Cryo-EM

Conference: 2020: 70th ACA Annual Meeting
08/03/2020: 12:00 PM  - 3:00 PM 
Oral Session 


Structural disorder and conformational flexibility in single particle cryo-EM

12:00 PM - 12:30 PM 
Single particle cryo-EM (cryo-electron microscopy) allows high-resolution imaging of macromolecular complexes in close-to-native state, at near-atomic resolutions. Cryo-EM has undergone several technological breakthroughs in microscopy, electron detectors, and image processing that have enabled its use recently in solving high-resolution structures of difficult proteins and complexes [1, 2]. As cryo-EM makes rapid progress, one of the key challenges that remains in the quest for higher resolutions on challenging targets is in dealing with conformational heterogeneity. Established methods for resolving discrete heterogeneity are ineffective in dealing with structural disorder and continuous conformational flexibility. Here, we describe two recent algorithmic developments in the cryoSPARC software system: non-uniform refinement for improving reconstructions in the presence of structural disorder, and 3D variability analysis for resolving continuous conformational motion and dynamics of protein molecules. We demonstrate that non-uniform refinement can improve the quality and resolution of 3D reconstructions for important targets such as membrane proteins, and that 3D variability analysis can capture significant aspects of molecular flexibility and conformational heterogeneity with the potential to reveal new biological insights into protein dynamics and function. [1] D Cressey and E Callaway, Nature 550 (2017), p. 167 [2] X Bai, G McMullan, and S Scheres, Trends in Biochemical Sciences 40 Issue 1 (2017), p. 49-57 

View Proposal 449


Ali Punjani, University of Toronto Toronto, ON 

Additional Author

David Fleet, University of Toronto Toronto, Ontario 

Single-particle cryo-EM at atomic resolution

12:30 PM - 1:00 PM 
The three-dimensional positions of atoms in protein molecules define their structure and the roles they perform in biological processes. The more precisely atomic coordinates are determined, the more chemical information can be derived and the more mechanistic insights into protein function may be inferred. With breakthroughs in electron detection and image processing technology, electron cryo-microscopy (cryo-EM) single-particle analysis has yielded protein structures with increasing levels of detail in recent years. However, obtaining cryo-EM reconstructions with sufficient resolution to visualise individual atoms in proteins has thus far been elusive. We recently demonstrated how a combination of new algorithms (Bayesian polishing, optical aberration correction and Ewald sphere correction) and hardware (a new cold field emission gun, new energy filter and the Falcon 4 detector) yields a 1.7 Å resolution cryo-EM reconstruction for a prototypical human membrane protein, the 3 GABAA receptor homopentamer. Such maps allow a detailed understanding of small molecule coordination, visualisation of solvent molecules and alternative conformations for multiple amino acids, as well as unambiguous building of ordered acidic side chains and glycans. Moreover, the scattering potential from many hydrogen atoms can be visualised in difference maps, allowing a direct analysis of hydrogen bonding networks. Applied to mouse apoferritin, our strategy led to a 1.2 Å resolution reconstruction that, for the first time, offers a genuine atomic resolution view of a protein molecule using single particle cryo-EM. 

View Proposal 446


Andrija Sente

Coffee Break

1:00 PM - 1:20 PM 

A complete data processing workflow for in situ structure determination with cryo-electron tomography

1:20 PM - 1:45 PM 
Electron cryotomography (CryoET) is currently the only method capable of visualizing cells in 3D at nanometer resolutions. While modern instruments produce massive amounts of tomography data containing extremely rich structural information, it takes tremendous effort to process those data and reach biological findings. Here we present an integrated workflow that covers the entire tomography data processing pipeline, from automated tilt series alignment and particle selection to subnanometer resolution subtomogram averaging. Resolution enhancement is made possible through the use of per-particle-per-tilt CTF correction and orientation refinement. This workflow greatly reduces human effort and increases the throughput of CryoET data processing, and is capable of determining protein structures at state-of-the-art resolutions for both purified macromolecules and cells. 

View Proposal 443


Muyuan Chen

New developments in Scipion 3.0: consolidating the project as a collaborative software framework.

1:45 PM - 2:10 PM 
Scipion has become a collaborative software framework where developers in different locations contribute to the project. One of its main advantages is that it provides Cryo-EM users with a unified platform where they can easily use and combine different software packages. With Scipion 3, we are consolidating the framework for future challenges and growing. We are in the process of completely migrating the project to Python 3, while improving the general design and performance of the system. At the same time, many new scientific tools have been added in the form of plugins. For example, several tomography programs will be available in the next release, as well as some tools for helical processing. Localized reconstruction tools have also been implemented as a plugin and have been enhanced in the new version. Additionally, some work is in progress to add some crystallography programs (DIALS and XDS). 

View Proposal 445


Jose Miguel de la Rosa Trevin

Neural network particle picking and denoising in cryoEM with Topaz

2:10 PM - 2:35 PM 
Single particle cryoEM projects are often hampered by low SNR particle views, which are missed by most particle pickers or severely de-prioritized causing junk to be preferentially picked. Moreover, for non-globular, small, asymmetric, and aggregated proteins, picking and centering such particles becomes critical. To solve these issues and more, we present Topaz particle picking using a novel positive-unlabelled framework and Topaz-Denoise using the Noise2Noise framework. We show that Topaz and Topaz-Denoise significantly increase the number of real particles picked, enable conventionally difficult projects, significantly decrease classification bias, and increase collection efficiency. We show the first in-depth analysis of pre-trained 2D and 3D denoising models for cryoEM and cryoET, which remove the characteristic sheets of noise in cryoEM micrographs of proteins and cryoET cellular tomograms. In the past few months, Topaz has been used by several labs around the world to enable and optimize single particle SARS-CoV2 projects. These recent projects and more will be highlighted to show the unique and timely power of Topaz. We will also highlight Topaz integration into the most popular cryoEM suites: Relion, CryoSPARC, Scipion, and Appion. 

View Proposal 447


Alex Noble, New York Structural Biology Center / Simons Electron Microscopy Center NEW YORK, NY 

Additional Author(s)

Tristan Bepler, MIT Cambridge, MA 
Bonnie Berger, MIT Cambridge, MA 

High-resolution in situ imaging of biological samples with Warp and M

2:35 PM - 3:00 PM 
In situ structural biology aims to capture the shape, interactions and organization of macromolecules inside their native cellular environment. Cryo-electron tomography offers an unparalleled level of detail for in situ studies. Instead of a single 2D projection, a series of images at different sample tilt angles (tilt series) is acquired. Once aligned, tomograms can be reconstructed from the tilt series to resolve the macromolecular organization in 3D. Tilt series data can also be subjected to single-particle analysis (SPA) to achieve even higher resolution by averaging the signal of multiple identical particles. Unfortunately, the resolution of SPA maps obtained from in situ data has trailed far behind the results achievable for isolated molecules in vitro. On the one hand, the crowded cellular environment and high sample thickness limit the available signal for each particle. On the other hand, accurate alignment of tomographic tilt series data poses a harder task than the 2D movies commonly acquired for in vitro samples, and is not yet handled optimally by the less mature software ecosystem. M is a new tool that integrates with Warp and RELION, and improves tilt series alignments by modeling the contents of each tomogram as a multi-particle system with hyperparameters describing its various modes of deformation. Individual particle poses are affected by the deformation model as a function of particle position, tilt angle and exposure. The spatial and temporal coarseness of the model regularizes the solution and prevents the overfitting of physically implausible changes. M fits particle poses and deformation hyperparameters simultaneously by maximizing the correlation between projections of high-resolution maps and experimental image data using a gradient descent optimization. Coupled with advances in high-defocus contrast transfer function correction and 3D map denoising, M outperforms other implementations of reference-based tilt series refinement. Most notably, M is able to refine in situ data of 70S ribosomes inside a 150 nm thick bacterium to 3.7 Å, rivaling the resolution of comparable in vitro studies for the first time. 

View Proposal 459


Dimitry Tegunov, Max Planck Institute for Biophysical Chemistry Goettingen