A Bayesian approach to extracting free-energy profiles from cryo-electron microscopy experiments.

Area of Science:

• Structural biology
• Biophysics
• Computational chemistry

Background:

• Cryo-electron microscopy (cryo-EM) provides high-resolution structural data of biomolecules.
• Individual cryo-EM particles offer potential insights into conformational variability and free-energy landscapes.
• Low signal-to-noise ratio (SNR) in single particles hinders the extraction of thermodynamic information.

Purpose of the Study:

• To develop a novel computational method for extracting free-energy profiles from cryo-EM data.
• To address the challenge of low SNR in single-particle cryo-EM for thermodynamic analysis.
• To enable a more comprehensive characterization of biomolecular conformational ensembles.

Main Methods:

• Introduction of the cryo-BIFE (cryo-EM Bayesian Inference of Free-Energy profiles) method.
• Utilizing a path collective variable to analyze cryo-EM images.
• Validation on synthetic systems with controlled imaging parameters and application to real cryo-EM data of a calcium-activated channel.

Main Results:

• The cryo-BIFE method successfully recovers free-energy profiles and their uncertainties from cryo-EM images.
• Pose accuracy and SNR were identified as critical factors for accurate free-energy recovery.
• The method identified both the most probable (calcium-bound) and a metastable (calcium-unbound) conformation of the calcium-activated channel.

Conclusions:

• The cryo-BIFE method enables the extraction of free-energy landscapes from cryo-EM data, overcoming SNR limitations.
• This approach allows for the characterization of conformational variability and thermodynamic ensembles of biomolecules.
• The study demonstrates the potential of cryo-EM as a single-molecule technique for detailed biophysical analysis.