Auto3D is a Python package for generating low-energy conformers from SMILES/SDF. It automatizes the stereoisomer enumeration and duplicate filtering process, 3D building process, fast geometry optimization and ranking process using ANI and AIMNet neural network atomistic potentials. Auto3D can be imported as a Python library, or be excuted from the terminal.

Please check out the information at documentation, including installation, usage, API and citation.

• Jupyter notebook examples can be found here
• To-do list for improvement and new features can be found here. You are welcomed to share your thoughts.
• Bugs go to the issues

1. Python >= 3.7
2. RDKit >= 2022.03.1(For the isomer engine)
3. PyTorch >= 2.1.0 (For the optimization engine)

If you have an environment with the above dependencies, Auto3D can be installed by

pip install Auto3D

Otherwise, you can create an environment and install Auto3D. In a terminal, the following code will create a environment named auto3D with Auto3D and its minimum dependencies installed.

git clone https://github.com/isayevlab/Auto3D_pkg.git
cd Auto3D_pkg
conda env create --file installation.yml --name auto3D
conda activate auto3D
pip install Auto3D