qmdesc

qmdesc is a package for QM atomic/bond descriptors prediction

At its core, qmdesc contains a derived ChemProp model that was trained on QM descriptors including hirshfeld partial charges, nucleophilicity/electrophilicity Fukui indcides, NMR shielding constants, bond lengths, and bond indices calculated under B3LYP/def2svp level of theory. This work was first presented in Regio-Selectivity Prediction with a Machine-Learned Reaction Representation and On-the-Fly Quantum Mechanical Descriptors.

qmdesc can be used through either python module or command line

Prerequisite: RDKit

Indices and tables