Related Experiment Video
Updated: Jul 26, 2025
13:43
A Robust Single-Particle Cryo-Electron Microscopy cryo-EM Processing Workflow with cryoSPARC, RELION, and Scipion
Published on: January 31, 2022
13.4K
A large expert-curated cryo-EM image dataset for machine learning protein particle picking
Ashwin Dhakal1, Rajan Gyawali1, Liguo Wang2
1Department of Electrical Engineering and Computer Science, NextGen Precision Health, University of Missouri, Columbia, MO, 65211, USA.
Scientific Data
|June 22, 2023
Summary
Cryo-electron microscopy (cryo-EM) particle picking is now faster with CryoPPP, a new expert-curated dataset. This large, labelled dataset aids automated protein structure determination using machine learning and AI.
Area of Science:
- Structural Biology
- Biophysics
- Computational Biology
Background:
- Cryo-electron microscopy (cryo-EM) is vital for determining biological macromolecular complex structures.
- Automating single-protein particle picking from cryo-EM images is essential for efficient structure reconstruction.
- Current template-based methods are laborious, and AI-driven approaches lack sufficient high-quality labelled training data.
Purpose of the Study:
- To introduce CryoPPP, a large, diverse, and expert-curated cryo-EM image dataset.
- To address the bottleneck of limited labelled data for automated particle picking in cryo-EM.
- To facilitate the development of advanced methods for automated protein particle picking.
Main Methods:
- Compiled 2.6 terabytes of labelled cryo-EM micrographs from 34 diverse protein datasets.
- Included 9,893 high-resolution micrographs with precisely labelled protein particle coordinates.
- Validated labelling accuracy using 2D particle class and 3D density map validation against gold standards.
Main Results:
- Developed CryoPPP, a comprehensive dataset for cryo-EM particle picking.
- The dataset provides extensive, validated labelled data for training and benchmarking.
- Demonstrated rigorous quality control through established validation protocols.
Conclusions:
- CryoPPP significantly advances the availability of labelled data for cryo-EM particle picking.
- The dataset will accelerate the development of both AI and classical automated particle picking methods.
- Expected to improve the efficiency and throughput of protein structure determination via cryo-EM.
