Due to the complexity and insufficient compute power alchemical free energy simulations are not routinely applied in early stage drug discovery. Another major issue is the unpredictability of the simulation stability and accuracy of the results. In the latest MOE release we added a streamlined interface to setup GPU enabled free energy calculations based on the Thermodynamic Integration (TI) method in AMBER18. The workflow includes structure preparation, ligand parameterization, simulation planning and analysis. Free energy calculations are known to be extremely sensitive and come with a high failure rate. For optimal simulation planning we designed a scoring function to minimize the expected error for a given system based on the available simulation resources (number of GPUs, nodes, etc.). Another major source of error is the simulation instability due to an imbalance between the electrostatic and van der Waals forces. Optimization of the AMBER soft-core potential resulted in much higher simulation stability and hence lower curvature of the potential function. We also suggest a new soft-core functional form which will further improve the stability. In conjunction with Fejér numerical integration the evaluation of this integral can be achieved with spectral accuracy. The workflow was validated through the calculation of absolute binding energies of 34 ligands of the p38 MAP Kinase and compared to experimental data.

Soft-core potential optimization to increase accuracy and stability in free energy calculations using AMBER thermodynamic integration in MOE

ACS Fall 2020 Virtual Meeting

The official archive and repository for all ACS Meetings & Expositions abstracts and select posters or presentations.