The performance of ANI-ML potentials for ligand-n(H2O) interaction energies and estimation of hydration free energies from end-point MD simulations

Temel M., Tayfuroglu O., Kocak A.

Journal of Computational Chemistry, vol.44, no.4, pp.559-569, 2023 (SCI-Expanded) identifier identifier identifier

  • Publication Type: Article / Article
  • Volume: 44 Issue: 4
  • Publication Date: 2023
  • Doi Number: 10.1002/jcc.27022
  • Journal Name: Journal of Computational Chemistry
  • Journal Indexes: Science Citation Index Expanded (SCI-EXPANDED), Scopus, Academic Search Premier, Applied Science & Technology Source, Chemical Abstracts Core, Chimica, Compendex, EMBASE, INSPEC, MEDLINE
  • Page Numbers: pp.559-569
  • Keywords: free energy, interaction energy, machine learning, molecular dynamics, solvation, LIGAND-BINDING, SOLVENT, COMPUTER
  • Kocaeli University Affiliated: No


© 2022 Wiley Periodicals LLC.Here, we investigate the performance of “Accurate NeurAl networK engINe for Molecular Energies” (ANI), trained on small organic compounds, on bulk systems including non-covalent interactions and applicability to estimate solvation (hydration) free energies using the interaction between the ligand and explicit solvent (water) from single-step MD simulations. The method is adopted from ANI using the Atomic Simulation Environment (ASE) and predicts the non-covalent interaction energies at the accuracy of wb97x/6-31G(d) level by a simple linear scaling for the conformations sampled by molecular dynamics (MD) simulations of ligand-n(H2O) systems. For the first time, we test ANI potentials' abilities to reproduce solvation free energies using linear interaction energy (LIE) formulism by modifying the original LIE equation. Our results on ~250 different complexes show that the method can be accurate and have a correlation of R2 = 0.88–0.89 (MAE <1.0 kcal/mol) to the experimental solvation free energies, outperforming current end-state methods. Moreover, it is competitive to other conventional free energy methods such as FEP and BAR with 15-20 × fold reduced computational cost.