Semi-Supervised JT-VAE for Molecular Graphs

Nov 2022: Our paper is accepted to AAAI 23' Deep Learning on Graphs workshop!

Usage

Quickstart

Run semi_jtvae_train.ipynb for training a model and run semi_jtvae_gen.ipynb for (conditionally) generating molecules using the trained model.

Citations

If you find the models useful in your research, we ask that you cite our paper:

@article{
  author={Atia Hamidizadeh, Tony Shen, Martin Ester},
  title={Semi-Supervised Junction Tree Variational Autoencoder for Molecular Graphs},
  year={2022},
  doi = {10.48550/ARXIV.2208.05119},
  url = {https://arxiv.org/abs/2208.05119},
}

License

This source code is licensed under the MIT license.