Simulations of water and hydrophobic hydration using a neural network potential
Using a neural network potential (ANI-1ccx) generated from quantum data on a large data set of molecules and pairs of molecules, isothermal, constant volume simulations demonstrate that the model can be as accurate as ab initio molecular dynamics for simulations of pure liquid water and the aqueous solvation of a methane molecule. No theoretical or experimental data for the liquid phase is used to train the model, suggesting that the ANI-1ccx approach is an effective method to link high level ab initio methods to potentials for large scale simulations.