Robust evaluation of 3D electron cryomicroscopy data using tilt-pairs.

Area of Science:

• * Structural Biology
• * Biophysics
• * Computational Biology

Background:

• * Electron microscopy (EM) is crucial for determining protein complex structures.
• * Low-dose imaging in EM generates noisy 2D particle images, complicating 3D map reconstruction.
• * Validation of reconstructed 3D density maps is essential due to potential alignment errors.

Purpose of the Study:

• * To establish a statistical framework for validating EM-derived 3D density maps.
• * To develop a quantitative metric for assessing the quality of particle images and maps.
• * To provide an objective method for verifying the correctness of reconstructed protein structures.

Main Methods:

• * Utilized tilt-pairs of electron micrographs with known angular relationships.
• * Modeled the directional distribution of angular data using a Fisher distribution on the unit sphere.
• * Introduced the concentration parameter (κ) as a metric for data and map quality.

Main Results:

• * Developed a quantitative metric, the concentration parameter κ, for evaluating EM datasets and density maps.
• * Demonstrated that a large κ indicates good agreement between particle images and the 3D density map.
• * Proposed recommendations for structure validation: κ > 10 and p-value < 0.01.

Conclusions:

• * The statistical framework offers an objective method to assess the correctness of reconstructed 3D density maps.
• * The concentration parameter κ provides a simple, comparable metric independent of specific EM data collection or analysis methods.
• * This approach enhances the reliability of structural information obtained from single-particle electron microscopy.