CryoSift - An accessible and automated CNN-driven tool for cryo-EM 2D class selection

Area of Science:

• Structural Biology
• Biophysics
• Computational Biology

Background:

• Single-particle cryo-electron microscopy (cryo-EM) is vital for determining molecular structures.
• Automating repetitive tasks in cryo-EM data processing is crucial for efficiency.
• Current methods often require significant manual intervention for particle selection.

Purpose of the Study:

• To develop an automated tool for assessing the quality of 2D averages in cryo-EM datasets.
• To enable automatic selection of suitable particles for high-resolution cryo-EM reconstructions.
• To integrate this tool into a fully automated processing pipeline.

Main Methods:

• Development of a platform-independent convolutional neural network (CNN) tool named CryoSift.
• Integration of CryoSift into an automated cryo-EM data processing pipeline using cryosparc-tools.
• Customizable 2D assessment workflow for particle selection.

Main Results:

• CryoSift effectively assesses the quality of 2D averages.
• Automated particle selection based on CryoSift's assessment improves reconstruction quality.
• The integrated pipeline facilitates high-throughput cryo-EM data processing.

Conclusions:

• CryoSift provides an automated solution for a key challenge in cryo-EM data processing.
• The tool enhances efficiency and accessibility for both novice and experienced cryo-EM users.
• This automated workflow supports high-resolution structure determination.