Advancing towards a "one-button" screening solution for Cryo-EM data acquisition with Smart EPU

For many years, the Cryo-EM community is expecting that software automation will achieve the ultimate goal of providing a Single Particle Analysis (SPA) data acquisition workflow that requires no setup from the user's side [1,2]. Achieving that will allow many more researchers, who are not necessarily application or microscope experts an optimal use of time in high throughput cryo-electron microscopes. This is particularly important in industrial set-ups or entry-level labs where efficiency and ease of use are key adoption barriers. Moreover, it has been identified that one of the main bottlenecks in Single Particle Acquisition (SPA) is screening automation [2]. Thermo Scientific Smart EPU with EPU Multigrid already provides a workflow that enables screening of multiple samples in highly automated way [3]. We are now expanding on that with the goal to make SPA screening fully unattended. With the use of EPU Multigrid, EPU Quality Monitor (EQM) and Embedded CryoSPARC Live, and a novel set of AI/DL algorithms, we can now demonstrate a workflow that allows users to screen a full cassette of samples on an autoloader system without the need to pass through any manual setup tasks. To achieve that level of automation, many parts of the workflow need to be optimized. User workflows for screening often include the acquisition and subsequent analysis of small datasets from numerous grids in different conditions to find optimal ice thickness and particle behavior [2]. This task is quite tedious: Users need to load each grid, select grid squares with a variation in ice thickness, identify foil holes, select foil holes with a variation in ice thickness, curate initial selections and afterwards perform data analysis to find favorable conditions for data acquisition. Smart EPU introduces a set of innovative AI/DL plugins that automate all the above steps. Experimental results show that the algorithms perform to a comparable level to an experienced user. We also combine these innovations with a redesigned UI to invoke the setup with just one click. This screening solution will be connected to EQM and Embedded CryoSPARC Live, so data analysis also happens in real time. With these developments in place, users will be able to load the samples, immediately start the screening process, and after a few hours get the results visualized in our data integration platform on their laptop.

Institutions

Publication Details

Journal
Structural Dynamics
Published
2025-09-01
DOI
https://doi.org/10.1063/4.0000878
Primary Topic
Advanced Electron Microscopy Techniques and Applications
Type
article
Field-Weighted Citation Impact
0.00
Controls
|||
ALL TIME
JAN
FEB
MAR
APR
MAY
JUN
JUL
AUG
SEP
OCT
article

Advancing towards a "one-button" screening solution for Cryo-EM data acquisition with Smart EPU

Structural Dynamics
Advanced Electron Microscopy Techniques and Applications
article

Advancing towards a "one-button" screening solution for Cryo-EM data acquisition with Smart EPU

article en

Abstract

For many years, the Cryo-EM community is expecting that software automation will achieve the ultimate goal of providing a Single Particle Analysis (SPA) data acquisition workflow that requires no setup from the user's side [1,2]. Achieving that will allow many more researchers, who are not necessarily application or microscope experts an optimal use of time in high throughput cryo-electron microscopes. This is particularly important in industrial set-ups or entry-level labs where efficiency and ease of use are key adoption barriers. Moreover, it has been identified that one of the main bottlenecks in Single Particle Acquisition (SPA) is screening automation [2]. Thermo Scientific Smart EPU with EPU Multigrid already provides a workflow that enables screening of multiple samples in highly automated way [3]. We are now expanding on that with the goal to make SPA screening fully unattended. With the use of EPU Multigrid, EPU Quality Monitor (EQM) and Embedded CryoSPARC Live, and a novel set of AI/DL algorithms, we can now demonstrate a workflow that allows users to screen a full cassette of samples on an autoloader system without the need to pass through any manual setup tasks. To achieve that level of automation, many parts of the workflow need to be optimized. User workflows for screening often include the acquisition and subsequent analysis of small datasets from numerous grids in different conditions to find optimal ice thickness and particle behavior [2]. This task is quite tedious: Users need to load each grid, select grid squares with a variation in ice thickness, identify foil holes, select foil holes with a variation in ice thickness, curate initial selections and afterwards perform data analysis to find favorable conditions for data acquisition. Smart EPU introduces a set of innovative AI/DL plugins that automate all the above steps. Experimental results show that the algorithms perform to a comparable level to an experienced user. We also combine these innovations with a redesigned UI to invoke the setup with just one click. This screening solution will be connected to EQM and Embedded CryoSPARC Live, so data analysis also happens in real time. With these developments in place, users will be able to load the samples, immediately start the screening process, and after a few hours get the results visualized in our data integration platform on their laptop.

Structural DynamicsVol. 12(5_Supplement)
Thermo Fisher Scientific (Netherlands) (NL), Thermo Fisher Scientific (United States) (US)
Industry, innovation and infrastructure
Openalex Percentile: Top 25%
Advanced Electron Microscopy Techniques and Applications
AI Navigator

Ask Laika to Summarize, Analyze, and Connect papers live on the map.

Summarize Papers & Methodologies

Extract key findings, datasets, and comparative methods across publications.

Benchmark Rankings & Visual Analytics

Rank top research institutions, authors, funders, topics, and journals by Field-Weighted Citation Impact (FWCI) and paper volume with instant charts.

Connect Distant Disciplines

Bridge topological clusters on the map to find hidden collaborative intersections.