Exploring the Effect of Enhanced Sampling on Protein Stability Prediction

Area of Science:

• Biochemistry
• Computational Biology
• Molecular Dynamics

Background:

• Protein stability is crucial for function.
• Predicting mutation effects on stability is a key challenge.
• Alchemical free-energy calculations offer a computational approach.

Purpose of the Study:

• To evaluate alchemical free-energy calculations for predicting protein stability changes.
• To assess the impact of sampling and starting structures on accuracy.
• To analyze mutation effects across proteins of varying complexity.

Main Methods:

• Classical molecular dynamics (MD) simulations in explicit solvent.
• Utilized the GROMOS force field.
• Performed alchemical free-energy calculations for 93 single-point mutations across three proteins.

Main Results:

• Calculations showed good agreement with experimental data for protein stability changes.
• Insufficient sampling in previous studies led to deviations.
• Starting structure dependence was significant for smaller proteins.

Conclusions:

• Alchemical free-energy calculations are effective for predicting mutation-induced stability changes.
• Hamiltonian replica exchange with adequate sampling and multiple starting structures ensures convergence.
• Accurate prediction requires careful consideration of simulation parameters and sampling depth.