About
AIdsorb is a Python package for deep learning on molecular point clouds.
AIdsorb aims to provide a simple, easy-to-use and reproduce interface for:
📥 Creating molecular point clouds
🤖 Training DL algorithms on molecular point clouds
Why molecular point clouds?
A molecular point cloud, being essentially a set of atomic positions, atomic numbers and any additional information, provides a natural and efficient way to represent molecular structures and chemical systems in a machine understandable format. This in turns allows us to perform DL directly on raw structural information, removing the need for manual feature extraction.
The above point cloud represents IRMOF-1. You can hover over the figure to play with it.
TODO
Enable users to make predictions from the command line.
Enable users to fine-tune models trained on large data.
This might require the usage of PyTorch Geometric.
Add more featurization options. These should be fast!
Contributing
🙌 We welcome contributions from the community to help improve and expand this project!
You can start by 🛠️ opening an issue for:
🐛 Reporting bugs
🌟 Suggesting new features
📚 Improving documentation
🎨 Adding your example to the Gallery
We appreciate your efforts to submit well-documented pull requests and participate in discussions.
💪 Together, we can make this project even better!
Citing
If you use AIdosrb in your research, please consider citing the following work:
Currently, not available.
License
AIdsorb is released under the GNU General Public License v3.0 only.