PACMAN-charge 1.1.0

Partial Atomic Charges for Porous Materials based on Graph Convolutional Neural Network (PACMAN)

PACMAN

A Partial Atomic Charge Predicter for Porous Materials based on Graph Convolutional Neural Network (PACMAN)

Usage

from PACMANCharge import pmcharge
pmcharge.predict(cif_file="./test/Cu-BTC.cif",charge_type="DDEC6",digits=6,atom_type=False,neutral=False)

• cif_file: cif file (without partial atomic charges) [cif path]
• charge-type (default: DDE6): DDEC6, Bader, CM5 or REPEAT
• digits (default: 6): number of decimal places to print for partial atomic charges. ML models were trained on a 6-digit dataset.
• atom-type (default: True): keep the same partial atomic charge for the same atom types (based on the similarity of partial atomic charges up to 2 decimal places).
• neutral (default: True): keep the net charge is zero. We use "mean" method to neuralize the system where the excess charges are equally distributed across all atoms.

Website & Zenodo

PACMAN-APPlink
github repositorylink
DOWNLOAD full code and datasetlink But we will not update new vesion in Zenodo.

Reference

If you use PACMAN Charge, please cite this paper:

@article{,
    title={PACMAN: A Robust Partial Atomic Charge Predicter for Nanoporous Materials based on Crystal Graph Convolution Network},
    author={Zhao, Guobin and Chung, Yongchul},
    journal={Journal of Chemical Theory and Computation},
    year={2024},
    DOI={10.1021/acs.jctc.4c00434}
}

Bugs

If you encounter any problem during using PACMAN, please email [e-mail].

Group: Molecular Thermodynamics & Advance Processes Laboratory