Deep Potential driven dihedral scan toolkit.
- Python 3.8 or later (3.10 recommended)
- PyTorch 2
- DeePMD-kit (at least 3.0.0a0)
- networkx
- RDKit
- ASE
- Geometric
- DPData
- OpenMM (for adding bias and reading Gromacs topology)
- tblite (for semi-empirical level geometry optimization)
- Psi4 (for high-level quantum mechanical single point energy)
Users can install the required packages by running the following command:
conda install -c conda-forge rdkit geometric tblite-python psi4 openmm
If you want to use Psi4 to compute RESP charges, you also need to install Psi4 with the following command:
conda install -c psi4 psi4
pip install ase dpdata networkx
If you want to use Gromacs topology, you also need to install ambertools, acpype and parmed.
conda install -c conda-forge ambertools
pip install acpype parmed
Please look at DeePMD-kit's installation guide
Accelerate Geometric using Numba and JIT. Please look at this repository.
Numba is also needed if you want to use the JIT version of Geometric.
conda install -c conda-forge numba
pip install .