Introducing sampling entropy in repository based adaptive umbrella sampling.

Area of Science:

• Computational Chemistry
• Statistical Mechanics
• Biophysics

Background:

• Determining free energy surfaces is crucial for understanding complex systems.
• High energy barriers and complex landscapes necessitate efficient simulation methods for uniform sampling.
• Existing methods require improvement in adaptivity and robustness.

Purpose of the Study:

• To introduce sampling entropy (SE) as a key metric for assessing uniform sampling and simulation convergence.
• To enhance the repository based adaptive umbrella sampling (RBAUS) method by integrating SE and the concentration theorem.
• To demonstrate the improved adaptivity, robustness, and applicability of the new RBAUS-SE approach.

Main Methods:

• Demonstrated the utility of sampling entropy (SE) for evaluating free energy simulations.
• Integrated SE and the concentration theorem into the biasing-potential-updating scheme of RBAUS.
• Applied the enhanced RBAUS-SE method to one-dimensional free energy profiles and two-dimensional free energy surfaces.

Main Results:

• Sampling entropy (SE) was confirmed as an excellent indicator for uniform sampling and simulation convergence.
• The RBAUS-SE approach exhibited improved adaptivity, robustness, and applicability compared to the original RBAUS.
• Successful determination of 1D free energy profiles and 2D free energy surfaces for alanine dipeptide in gas and aqueous phases.

Conclusions:

• The RBAUS-SE method provides a more efficient and reliable approach for calculating free energy surfaces.
• Sampling entropy is a valuable tool for monitoring and ensuring the quality of free energy simulations.
• The enhanced method is broadly applicable to various systems, including complex molecular simulations.