Intergrates PINN, E3NN and data augmentation for induced dipole prediction

Requirements

• Conda
• OpenMM MPIDPlugin
• Cuda-enabled GPU

Environment setup

cd <PATH_TO_pid>
conda env create -n=<ENV_NAME> -f env.yaml
conda activate <ENV_NAME>
pip install -e .

Unittest

• pytest

Usage

Please check examples pid/examples.
python pid/app/pinn.py pid/examples/water_12A_10.pdb --config pid/examples/pid_example_config.yaml --tensor-database pid/examples/water_12A_10.db
For detailed configuration in both optional arguments and config file, please use
python pid/app/pinn.py -h