Combining Force Fields and Neural Networks for an Accurate Representation of Bonded Interactions

We present a formalism of a neural network encoding of bonded interactions in molecules. This intramolecular encoding is consistent with models of intermolecular interactions previously designed by this group. Variants of the encoding fed into a corresponding neural network may be used to economically improve representation of torsional degrees of freedom in any force field. We test the accuracy of the reproduction of the ab initio potential energy surface on a set of conformations of two dipeptides - methyl-capped ALA and ASP - in several scenarios. The encoding, either alone or in conjunction with an analytical potential, improves agreement with ab initio energies on par with that of other neural network-based potentials. Using the encoding and neural nets in tandem with an analytical model places the agreements firmly within ‘chemical accuracy’ of ± 0.5 kcal/mol.