qmdesc

A trained multitask constraint message passing neural networks for QM atomic/bond property predictions as described in the paper Regio-Selectivity Prediction with a Machine-Learned Reaction Representation and On-the-Fly Quantum Mechanical Descriptors.

QM descriptors under B3LYP/def2svp level of theory that can be predicted with this model:

1. Hirshfeld partial charge
2. Neucleuphilic Fukui indices
3. Electrophilic Fukui indices
4. NMR shielding constants
5. Bond lengths
6. Bond orders

Documentation: Documentation of qmdesc is available at https://qmdesc.readthedocs.io/en/latest/index.html.

Requirements

• RDKit

Installation

For all installations, we recommend using conda to get the necessary rdkit dependency:

conda install -c rdkit rdkit
pip install qmdesc

Or from envrioment.yml

conda create --name qmdesc --file environment.yml