This repository contains code to generate and analyze molecular conformers. The code uses RDKit to create conformers from SMILES, and returns a list of ASE Atoms. Optionally, the code allows optimizing them at the CCSD(T)-level with a pre-trained neural network potential (ANI-1ccx implemented in TorchANI).
To use this package, first clone the repository and install it under your virtual environment using pip:
git clone https://github.com/digital-synthesis-lab/conformers.git
cd conformers
pip install -e .
The conformers package contains an API-like interface that can be imported in other code, and scripts that enable computing the conformers using the command line interface.
These scripts can be accessed using the conformers command that is installed with the repository.
For example, to generate an ASE Atoms object based on a single conformer:
from conformers.confgen import ConformerGenerator
from conformers.convert import mol_to_atoms
cgen = ConformerGenerator(smiles="CN1C=NC2=C1C(=O)N(C(=O)N2C)C")
mol, energies = cgen.run()
atoms = mol_to_atoms(mol)