FSC-Q: a CryoEM map-to-atomic model quality validation based on the local Fourier shell correlation

Area of Science:

• Structural Biology
• Biophysics
• Computational Biology

Background:

• Advances in cryo-electron microscopy (cryoEM) have led to a surge in solved atomic models.
• The quality of cryoEM maps is known to vary locally, necessitating local evaluation of derived structural models.

Purpose of the Study:

• To present a novel method for the local analysis of map-to-model fit in cryoEM.
• To introduce a quantitative measure, FSC-Q, for assessing how well atomic models are supported by cryoEM map signal.

Main Methods:

• Developed an algorithm that compares two local resolution maps.
• The first local resolution map is the Fourier shell correlation (FSC) between the full cryoEM map and the atomic model.
• The second local resolution map is calculated between the half maps used in single particle analysis.

Main Results:

• Introduced FSC-Q as a quantitative estimation of model support by map signal content.
• Demonstrated that FSC-Q can effectively detect overfitting in cryoEM structural models.
• Showcased FSC-Q as a valuable complement to existing methods like the Q-score for atomic resolvability assessment.

Conclusions:

• FSC-Q provides a robust method for local evaluation of cryoEM map-to-model fit.
• This quantitative measure aids in validating structural models and identifying potential overfitting issues.
• FSC-Q enhances the reliability of atomic models derived from cryoEM data.