An anisotropic message-passing neural network, embedded electrostatically in a classical MM environment, predicts solution-phase reaction free energies in explicit solvent with near-chemical accuracy in three benchmark applications.
Title resolution pending
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
physics.chem-ph 1years
2024 1verdicts
CONDITIONAL 1representative citing papers
citing papers explorer
-
Neural Network Potential with Multi-Resolution Approach Enables Accurate Prediction of Reaction Free Energies in Solution
An anisotropic message-passing neural network, embedded electrostatically in a classical MM environment, predicts solution-phase reaction free energies in explicit solvent with near-chemical accuracy in three benchmark applications.