A first-principles exploration of the conformational space of sodiated pyranose assisted by neural network potentials
Huu Trong Phan, Pei‐Kang Tsou, Po‐Jen Hsu, Jer‐Lai Kuo
Abstract
An NNP created by the active learning scheme was used to assist the exploration of the conformational space of sodiated pyranoses. As a result, more than 17 000 distinct local minima at the DFT level were located and an NNP with an accuracy of 1 kJ mol −1 was created.
Topics & Concepts
Maxima and minimaConformational isomerismChemistryDissociation (chemistry)PyranoseArtificial neural networkComputational chemistryBiological systemComputer scienceArtificial intelligenceMoleculePhysical chemistryMathematicsStereochemistryBiologyMathematical analysisOrganic chemistryComputational Drug Discovery MethodsProtein Structure and DynamicsMolecular spectroscopy and chirality