Junction tree variational autoencoder for molecular graph generation
© 2018 by authors.All right reserved. 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...
| Main Authors: | , , |
|---|---|
| Other Authors: | |
| Format: | Article |
| Language: | English |
| Published: |
2021
|
| Online Access: | https://hdl.handle.net/1721.1/137391 |
| _version_ | 1826216973795590144 |
|---|---|
| author | Jin, Wengong Barzilay, Regina Jaakkola, Tommi |
| author2 | Massachusetts Institute of Technology. Computer Science and Artificial Intelligence Laboratory |
| author_facet | Massachusetts Institute of Technology. Computer Science and Artificial Intelligence Laboratory Jin, Wengong Barzilay, Regina Jaakkola, Tommi |
| author_sort | Jin, Wengong |
| collection | MIT |
| description | © 2018 by authors.All right reserved. 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. |
| first_indexed | 2024-09-23T16:56:28Z |
| format | Article |
| id | mit-1721.1/137391 |
| institution | Massachusetts Institute of Technology |
| language | English |
| last_indexed | 2024-09-23T16:56:28Z |
| publishDate | 2021 |
| record_format | dspace |
| spelling | mit-1721.1/1373912022-10-03T09:15:51Z Junction tree variational autoencoder for molecular graph generation Jin, Wengong Barzilay, Regina Jaakkola, Tommi Massachusetts Institute of Technology. Computer Science and Artificial Intelligence Laboratory © 2018 by authors.All right reserved. 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. 2021-11-04T18:58:53Z 2021-11-04T18:58:53Z 2018 2019-05-07T18:07:50Z Article http://purl.org/eprint/type/ConferencePaper https://hdl.handle.net/1721.1/137391 Jin, Wengong, Barzilay, Regina and Jaakkola, Tommi. 2018. "Junction tree variational autoencoder for molecular graph generation." en Creative Commons Attribution-Noncommercial-Share Alike http://creativecommons.org/licenses/by-nc-sa/4.0/ application/pdf arXiv |
| spellingShingle | Jin, Wengong Barzilay, Regina Jaakkola, Tommi Junction tree variational autoencoder for molecular graph generation |
| title | Junction tree variational autoencoder for molecular graph generation |
| title_full | Junction tree variational autoencoder for molecular graph generation |
| title_fullStr | Junction tree variational autoencoder for molecular graph generation |
| title_full_unstemmed | Junction tree variational autoencoder for molecular graph generation |
| title_short | Junction tree variational autoencoder for molecular graph generation |
| title_sort | junction tree variational autoencoder for molecular graph generation |
| url | https://hdl.handle.net/1721.1/137391 |
| work_keys_str_mv | AT jinwengong junctiontreevariationalautoencoderformoleculargraphgeneration AT barzilayregina junctiontreevariationalautoencoderformoleculargraphgeneration AT jaakkolatommi junctiontreevariationalautoencoderformoleculargraphgeneration |