A two step approach for semi-automated particle selection from low contrast cryo-electron micrographs

Area of Science:

• Structural Biology
• Biophysics
• Biochemistry

Background:

• Cryo-electron microscopy (cryo-EM) advances enable routine high-resolution macromolecular structure determination.
• Collecting large datasets (>10,000 images) is necessary for high-resolution cryo-EM.
• Manual particle selection is a bottleneck, especially for low-contrast, high-noise images of small molecules.

Purpose of the Study:

• To develop a general automated particle picking method for cryo-EM.
• To apply and validate the method on challenging micrographs of Escherichia coli RNA polymerase.
• To compare cryo-EM reconstructions from manually and automatically picked data.

Main Methods:

• A two-step automated particle picking approach was developed.
• The method was applied to micrographs of Escherichia coli RNA polymerase.
• Independent cryo-EM reconstructions were performed using manual and automated particle selection.

Main Results:

• The automated method successfully picked particles from low-contrast, high-noise micrographs.
• Cryo-EM reconstructions from manual and automated data were comparable.
• The reconstruction using automatically picked data achieved a higher resolution.

Conclusions:

• The developed automated method is general and effective for challenging cryo-EM datasets.
• Automated particle picking overcomes limitations of manual selection, improving efficiency and resolution.
• This approach facilitates high-resolution structural studies of small molecules using cryo-EM.