Insights and Challenges in Correcting Force Field Based Solvation Free Energies Using A Neural Network Potential

We present a comprehensive study investigating the potential gain in accuracy for calculating absolute solvation free energies (ASFE) using a neural network potential for the intramolecular energies. We calculated ASFE using the Open Force Field (OpenFF) and compared the results to previously calculated ASFEs employing the CHARMM General Force Field (CGenFF). By applying a nonequilibrium (NEQ) switching approach between the molecular mechanics (MM) description (either OpenFF or CGenFF) and the machine learning (ML)/MM level of theory (using ANI-2x as the ML potential), we attempt to improve the accuracy of the calculated ASFEs. The predictive performance of the results did not change when applying this approach to all 589 small molecules in the FreeSolv database that ANI-2x can describe. When selecting a subset of 156 molecules, focusing on compounds where the force fields performed poorly, we saw a slight improvement in the root-mean-square error (RMSE) and mean absolute error (MAE). The majority of our calculations utilized unidirectional NEQ protocols based on Jarzynski's equation. Additionally, we conducted bidirectional NEQ switching for the subset of 156 FreeSolv molecules. Notably, only a small fraction (10 out of 156) exhibited statistically significant discrepancies between unidirectional and bidirectional NEQ switching free energy estimates.