Thresholding of cryo-EM density maps by false discovery rate control

Area of Science:

• Structural Biology
• Biophysics
• Computational Biology

Background:

• Cryo-electron microscopy (cryo-EM) generates high-resolution structural data for macromolecules.
• Interpreting cryo-EM maps is challenging due to signal limitations and thresholding difficulties.
• Accurate interpretation is crucial for understanding molecular mechanisms.

Purpose of the Study:

• To develop a statistical framework for generating confidence maps from cryo-EM data.
• To improve the interpretation of cryo-EM density maps, especially at lower resolutions.
• To provide a reliable method for distinguishing signal from noise in cryo-EM maps.

Main Methods:

• Utilized multiple hypothesis testing and false discovery rate (FDR) control.
• Developed a statistical framework to generate three-dimensional confidence maps.
• Applied confidence maps and FDR-based thresholding to analyze cryo-EM data.

Main Results:

• Generated confidence maps that separate signal from background noise.
• Demonstrated the utility of confidence maps for near-atomic and lower-resolution cryo-EM maps.
• Showcased confidence maps as a conservative interpretation tool with minimized noise.

Conclusions:

• Confidence maps offer a robust method for interpreting cryo-EM density maps.
• This approach aids in analyzing structures with conformational flexibility or low occupancy.
• The framework provides a reliable measure of detection error against background noise.