Peptide Conformation Analysis Using an Integrated Bayesian Approach.

Area of Science:

• Protein biophysics
• Computational biology
• Structural biology

Background:

• Unfolded proteins exist as dynamic ensembles, unlike structured native proteins.
• Defining these ensembles is crucial for understanding protein folding and intrinsically disordered proteins.
• Short peptides, like alanine oligomers, serve as models for peptide backbone conformational preferences.

Purpose of the Study:

• To develop and apply an integrated computational-experimental-Bayesian approach to quantify the conformational ensembles of unfolded peptides.
• To objectively determine peptide conformational substates using molecular dynamics and clustering.
• To refine conformational ensemble estimates using experimental data and Bayesian analysis.

Main Methods:

• Molecular dynamics simulations to generate peptide conformational snapshots.
• Objective clustering of snapshots based on structural and dynamic features.
• Computation of spectroscopic data for each substate.
• Bayesian statistical analysis integrating experimental (NMR coupling constants) and computed data.

Main Results:

• An objective method for identifying and characterizing peptide conformational substates was demonstrated.
• The approach provides an estimate of substate populations with confidence intervals.
• The method was successfully applied to characterize the conformational ensemble of trialanine and trivaline in water.

Conclusions:

• The integrated approach effectively quantifies the conformational ensembles of unfolded peptides.
• This method offers a robust framework for refining conformational distributions with additional data.
• The findings advance the biophysical understanding of peptide backbone dynamics and intrinsically disordered proteins.