Cryo-EM single-particle structure refinement and map calculation using Servalcat.

Area of Science:

• Structural Biology
• Biophysics
• Computational Biology

Background:

• Cryo-electron microscopy (cryo-EM) single-particle analysis has reached atomic resolution.
• Accurate determination of atomic coordinates is crucial for high-resolution cryo-EM reconstructions.
• Existing tools may not fully support the detailed analysis required for atomic model refinement.

Purpose of the Study:

• To introduce Servalcat, a novel Python package for atomic model refinement in cryo-EM.
• To enhance the accuracy and detail of atomic models derived from cryo-EM data.
• To facilitate the visualization of weak electron density features, including hydrogen atoms.

Main Methods:

• Servalcat implements an atomic model refinement pipeline using REFMAC5 from the CCP4 package.
• It calculates weighted F_o - F_c difference maps using Bayesian statistics.
• The package supports atomic model refinement under symmetry constraints.

Main Results:

• Servalcat facilitates the refinement of atomic coordinates in cryo-EM reconstructions.
• The generated F_o - F_c maps aid in manual and automatic model building in real space.
• Weak features, such as hydrogen atom densities, become more visible, even at resolutions around 1.8 Å.

Conclusions:

• Servalcat is a valuable tool for improving atomic model accuracy in cryo-EM.
• The ability to visualize hydrogen densities enhances structural analysis.
• Symmetry constraint refinement improves model building for symmetric reconstructions.