Molecular dynamics-based refinement and validation for sub-5 Å cryo-electron microscopy maps.

Area of Science:

• Structural biology
• Biophysics
• Computational biology

Background:

• Cryo-electron microscopy (EM) provides low-resolution maps, requiring robust methods for accurate structure determination.
• Molecular dynamics flexible fitting (MDFF) is a powerful paradigm for fitting atomic models into EM density maps.

Purpose of the Study:

• To present two novel MDFF methods for resolving sub-5 Å cryo-EM maps.
• To enable accurate modeling of single structures and structural ensembles.
• To provide map-model validation and B-factor determination.

Main Methods:

• Cascade MDFF: Sequentially refines a search model against maps of increasing resolution.
• Resolution exchange MDFF: Iteratively refines models using maps of varying resolutions.
• Utilizes local root mean square fluctuations to capture macromolecular dynamics.

Main Results:

• Accurate modeling of β-galactosidase (3.2 Å) and TRPV1 (3.4 Å) proteins was achieved.
• Demonstrated a radius of convergence of approximately 25 Å.
• Methods successfully resolve both single structures and ensembles from cryo-EM maps.

Conclusions:

• The presented MDFF methods enhance the accuracy of cryo-EM structure determination.
• These approaches offer inherent validation and insights into macromolecular dynamics.
• Accessible cloud-based tools are available for researchers.