Junction Tree Variational Autoencoder for Molecular Graph Generation

Wengong Jin, Regina Barzilay, Tommi Jaakkola
Proceedings of the 35th International Conference on Machine Learning, PMLR 80:2323-2332, 2018.

Abstract

We seek to automate the design of molecules based on specific chemical properties. In computational terms, this task involves continuous embedding and generation of molecular graphs. Our primary contribution is the direct realization of molecular graphs, a task previously approached by generating linear SMILES strings instead of graphs. Our junction tree variational autoencoder generates molecular graphs in two phases, by first generating a tree-structured scaffold over chemical substructures, and then combining them into a molecule with a graph message passing network. This approach allows us to incrementally expand molecules while maintaining chemical validity at every step. We evaluate our model on multiple tasks ranging from molecular generation to optimization. Across these tasks, our model outperforms previous state-of-the-art baselines by a significant margin.

Cite this Paper

BibTeX
@InProceedings{pmlr-v80-jin18a, title = {Junction Tree Variational Autoencoder for Molecular Graph Generation}, author = {Jin, Wengong and Barzilay, Regina and Jaakkola, Tommi}, booktitle = {Proceedings of the 35th International Conference on Machine Learning}, pages = {2323--2332}, year = {2018}, editor = {Dy, Jennifer and Krause, Andreas}, volume = {80}, series = {Proceedings of Machine Learning Research}, month = {10--15 Jul}, publisher = {PMLR}, pdf = {http://proceedings.mlr.press/v80/jin18a/jin18a.pdf}, url = {https://proceedings.mlr.press/v80/jin18a.html}, abstract = {We seek to automate the design of molecules based on specific chemical properties. In computational terms, this task involves continuous embedding and generation of molecular graphs. Our primary contribution is the direct realization of molecular graphs, a task previously approached by generating linear SMILES strings instead of graphs. Our junction tree variational autoencoder generates molecular graphs in two phases, by first generating a tree-structured scaffold over chemical substructures, and then combining them into a molecule with a graph message passing network. This approach allows us to incrementally expand molecules while maintaining chemical validity at every step. We evaluate our model on multiple tasks ranging from molecular generation to optimization. Across these tasks, our model outperforms previous state-of-the-art baselines by a significant margin.} }
Endnote
%0 Conference Paper %T Junction Tree Variational Autoencoder for Molecular Graph Generation %A Wengong Jin %A Regina Barzilay %A Tommi Jaakkola %B Proceedings of the 35th International Conference on Machine Learning %C Proceedings of Machine Learning Research %D 2018 %E Jennifer Dy %E Andreas Krause %F pmlr-v80-jin18a %I PMLR %P 2323--2332 %U https://proceedings.mlr.press/v80/jin18a.html %V 80 %X We seek to automate the design of molecules based on specific chemical properties. In computational terms, this task involves continuous embedding and generation of molecular graphs. Our primary contribution is the direct realization of molecular graphs, a task previously approached by generating linear SMILES strings instead of graphs. Our junction tree variational autoencoder generates molecular graphs in two phases, by first generating a tree-structured scaffold over chemical substructures, and then combining them into a molecule with a graph message passing network. This approach allows us to incrementally expand molecules while maintaining chemical validity at every step. We evaluate our model on multiple tasks ranging from molecular generation to optimization. Across these tasks, our model outperforms previous state-of-the-art baselines by a significant margin.
APA
Jin, W., Barzilay, R. & Jaakkola, T.. (2018). Junction Tree Variational Autoencoder for Molecular Graph Generation. Proceedings of the 35th International Conference on Machine Learning, in Proceedings of Machine Learning Research 80:2323-2332 Available from https://proceedings.mlr.press/v80/jin18a.html.

Related Material