Biasing Potential Replica Exchange Multisite λ-Dynamics for Efficient Free Energy Calculations.

Area of Science:

• Computational Chemistry
• Molecular Dynamics
• Free Energy Calculations

Background:

• Traditional free energy methods struggle with scalability and convergence for large perturbations.
• Alchemical free energy calculations are crucial for drug discovery but often computationally expensive.

Purpose of the Study:

• To develop and validate a novel, efficient, and scalable free energy calculation method.
• To improve sampling and accuracy in alchemical free energy transformations, especially for flexible molecules.

Main Methods:

• Development of biasing potential replica exchange multisite λ-dynamics (BP-REX MSλD).
• Application of BP-REX MSλD and traditional MSλD to 2,5-benzoquinone derivatives.
• Optimization of sampling parameters (c) and development of a fragment transition protocol.
• Testing on geldanamycin-based Hsp90 inhibitors with significant volume changes.

Main Results:

• BP-REX MSλD demonstrated a 2-5 fold increase in transitions compared to traditional MSλD.
• Optimized protocols improved sampling efficiency for flexible moieties.
• Accurate prediction of binding free energies for Hsp90 inhibitors (AUE of 1.5 kcal/mol).

Conclusions:

• BP-REX MSλD is a highly efficient and scalable free energy method.
• The method enables routine calculations for hundreds of compounds with minimal simulations.
• BP-REX MSλD significantly advances computational drug discovery capabilities.