Simple, Efficient, and Reliable Computation of Multiple Free Energy Differences from a Single Simulation: A Reference...

Area of Science:

• Computational Chemistry
• Statistical Mechanics
• Molecular Dynamics

Background:

• Enveloping Distribution Sampling (EDS) is a powerful technique for calculating free energy differences.
• Optimization of reference Hamiltonian parameters in EDS can be computationally intensive and requires user expertise.
• Accurate free energy calculations are crucial for understanding chemical processes and designing new molecules.

Purpose of the Study:

• To present a novel automatic adaptive scheme for rapid optimization of reference Hamiltonian parameters in EDS.
• To evaluate the performance of six variants of the proposed update scheme.
• To demonstrate the scheme's applicability to complex condensed-phase systems and its ease of use for future studies.

Main Methods:

• Development and implementation of an automatic adaptive scheme for EDS parameter optimization.
• Testing of six different update scheme variants on a condensed-phase system involving water molecule insertion/deletion.
• Analysis of free energy estimates and configurational sampling regions.

Main Results:

• All six tested schemes yielded accurate free energy estimates.
• Absolute errors were as low as 0.1 kJ/mol for the best scheme and up to 1 kJ/mol for the worst.
• The scheme focuses configurational sampling on critical energy distribution intersection regions, ensuring high accuracy and precision.

Conclusions:

• The new automatic adaptive scheme significantly simplifies the application of EDS by eliminating the need for user-defined reference state Hamiltonian parameters.
• This method enables straightforward application of EDS to diverse systems, including ligand binding studies.
• The scheme offers a robust and efficient approach for high-accuracy free energy calculations.