Adaptive Lambda Scheduling: A Method for Computational Efficiency in Free Energy Perturbation Simulations.

Area of Science:

• Computational chemistry
• Molecular modeling
• Drug discovery

Background:

• Ligand-protein relative binding free energy (RBFE) calculations are crucial for drug discovery.
• Increased computational power has improved RBFE accessibility, but calculations remain resource-intensive.
• Optimizing the transformation coordinate lambda (λ) can reduce computational effort.

Purpose of the Study:

• To introduce Adaptive Lambda Scheduling (ALS), an efficient method for optimizing λ scheduling.
• To demonstrate ALS's ability to reduce computational costs in RBFE calculations.
• To validate that ALS maintains the predictive performance of RBFE calculations.

Main Methods:

• Developed Adaptive Lambda Scheduling (ALS) for on-the-fly, bespoke λ scheduling.
• Applied ALS to RBFE calculations.
• Evaluated computational cost and predictive performance compared to traditional methods.

Main Results:

• ALS achieves substantial reductions in computational cost for RBFE calculations.
• The method retains high predictive performance, comparable to existing approaches.
• ALS offers a streamlined and efficient approach to RBFE optimization.

Conclusions:

• Adaptive Lambda Scheduling (ALS) is an effective strategy for reducing computational demands in RBFE calculations.
• ALS provides a practical solution for making RBFE calculations more accessible in drug discovery.
• This approach enhances the efficiency of computational chemistry workflows.