Umbrella Sampling MD Simulations for Retention Prediction in Peptide Reversed-phase Liquid Chromatography.

Area of Science:

• Analytical Chemistry
• Computational Chemistry
• Proteomics

Background:

• Reversed-phase liquid chromatography (RPLC) is vital for separating complex peptide mixtures in bottom-up proteomics.
• Accurate prediction of peptide retention behavior in RPLC is of significant interest.
• Existing prediction methods rely on empirical rules or extensive training data.

Purpose of the Study:

• To explore a novel, first-principles modeling strategy for predicting peptide RPLC retention.
• To investigate the utility of molecular dynamics (MD) simulations for RPLC retention prediction.
• To establish a correlation between peptide-stationary phase interactions and experimental retention times.

Main Methods:

• Utilized umbrella sampling MD simulations to model peptide interactions with a C18 stationary phase.
• Determined the binding free energy (ΔGbinding) as a function of peptide-stationary phase distance.
• Compared simulation-derived ΔGbinding values with experimental retention times of tryptic peptides.

Main Results:

• Conventional MD simulations showed poor correlation between 'fraction bound' and experimental retention times.
• Umbrella sampling MD successfully determined ΔGbinding values under various mobile phase conditions.
• A linear correlation was observed between ΔGbinding and experimental RPLC retention times for test peptides.

Conclusions:

• Umbrella sampling MD provides a robust, first-principles approach for predicting peptide RPLC retention.
• This method bypasses the need for RPLC-specific reference data, enabling predictions for novel phases and analytes.
• The approach can be used independently or to enhance existing retention prediction algorithms.