Bayesian inference of conformational state populations from computational models and sparse experimental observables

Area of Science:

• Computational chemistry
• Molecular modeling
• Biophysics

Background:

• Estimating molecular conformational states is crucial for understanding their function.
• Existing methods often struggle with small molecules or require extensive experimental data.
• High-resolution structural models are essential for accurate conformational analysis.

Purpose of the Study:

• To present a novel Bayesian inference approach for determining conformational state populations.
• To apply the method to small molecules using high-resolution structural models.
• To validate the approach using sparse experimental data and assess its performance against existing methods.

Main Methods:

• Bayesian inference framework combined with molecular modeling.
• Inferential structure determination with reference potentials.
• Markov Chain Monte Carlo (MCMC) for sampling conformational states.
• Application to cineromycin B using Nuclear Magnetic Resonance (NMR) data and QM-refined ensembles.

Main Results:

• Accurate estimation of solution-state conformational populations for cineromycin B.
• Improved agreement with experimental data compared to previous modeling efforts.
• Quantification of computational modeling consistency with experimental data using Bayes factors.

Conclusions:

• The developed Bayesian approach effectively estimates conformational populations for small molecules.
• The method integrates molecular modeling and sparse experimental data for enhanced accuracy.
• It provides a robust framework for analyzing molecular conformations and validating computational models.