Geometry Optimization with Machine Trained Topological Atoms

• Published in: Scientific reports 2017-10, Vol.7 (1), p.12817-18, Article 12817
• Main Authors: Zielinski, François, Maxwell, Peter I., Fletcher, Timothy L., Davie, Stuart J., Di Pasquale, Nicodemo, Cardamone, Salvatore, Mills, Matthew J. L., Popelier, Paul L. A.
• Format: Article
• Language: English
• Subjects: 119/118; 639/638/563/606; 639/638/563/758; Energy; Geometry; Humanities and Social Sciences; INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CHEMISTRY; Learning algorithms; multidisciplinary; Potential energy; Science; Science & Technology - Other Topics; Science (multidisciplinary)
• ISSN: 2045-2322; 2045-2322
• DOI: 10.1038/s41598-017-12600-3

The geometry optimization of a water molecule with a novel type of energy function called FFLUX is presented, which bypasses the traditional bonded potentials. Instead, topologically-partitioned atomic energies are trained by the machine learning method kriging to predict their IQA atomic energies for a previously unseen molecular geometry. Proof-of-concept that FFLUX’s architecture is suitable for geometry optimization is rigorously demonstrated. It is found that accurate kriging models can optimize 2000 distorted geometries to within 0.28 kJ mol −1 of the corresponding ab initio energy, and 50% of those to within 0.05 kJ mol −1 . Kriging models are robust enough to optimize the molecular geometry to sub-noise accuracy, when two thirds of the geometric inputs are outside the training range of that model. Finally, the individual components of the potential energy are analyzed, and chemical intuition is reflected in the independent behavior of the three energy terms E intra A (intra-atomic), V cl AA ' (electrostatic) and V x AA ' (exchange), in contrast to standard force fields.