Bayesian refinement of protein structures and ensembles against SAXS data using molecular dynamics.

Area of Science:

• Structural biology
• Biophysics
• Computational biology

Background:

• Small-angle X-ray scattering (SAXS) is vital for studying protein structures and ensembles in solution.
• SAXS data refinement is often ambiguous due to low information content and unknown systematic errors.
• Accurate structural determination requires addressing limitations in SAXS data analysis.

Purpose of the Study:

• To develop a robust method for refining protein structures and ensembles using SAXS data.
• To overcome ambiguities in SAXS data analysis by integrating Bayesian inference and molecular dynamics.
• To accurately model protein conformational heterogeneity and account for experimental uncertainties.

Main Methods:

• Integration of Bayesian inference with all-atom molecular dynamics simulations.
• Explicit-solvent small-angle X-ray scattering (SAXS) calculations.
• Probabilistic assessment of the number of conformational states required for data fitting.

Main Results:

• The Bayesian approach effectively weights SAXS data against prior physical knowledge.
• Quantification of structural precision and ambiguity in fitted ensembles.
• Successful validation using a periplasmic binding protein and application to heat shock protein 90 (Hsp90).

Conclusions:

• The developed method enhances the accuracy and reliability of SAXS data-driven structural studies.
• Apo Hsp90 is consistent with a single wide-open conformation.
• ATP-bound Hsp90 suggests a heterogeneous ensemble of closed and wide-open states, revealing conformational dynamics.