Towards efficient discovery of green synthetic pathways with Monte Carlo tree search and reinforcement learning
© The Royal Society of Chemistry. Computer aided synthesis planning of synthetic pathways with green process conditions has become of increasing importance in organic chemistry, but the large search space inherent in synthesis planning and the difficulty in predicting reaction conditions make it a s...
| Main Authors: | , , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Royal Society of Chemistry (RSC)
2021
|
| Online Access: | https://hdl.handle.net/1721.1/135198 |
| _version_ | 1826214297565396992 |
|---|---|
| author | Wang, Xiaoxue Qian, Yujie Gao, Hanyu Coley, Connor W Mo, Yiming Barzilay, Regina Jensen, Klavs F |
| author_facet | Wang, Xiaoxue Qian, Yujie Gao, Hanyu Coley, Connor W Mo, Yiming Barzilay, Regina Jensen, Klavs F |
| author_sort | Wang, Xiaoxue |
| collection | MIT |
| description | © The Royal Society of Chemistry. Computer aided synthesis planning of synthetic pathways with green process conditions has become of increasing importance in organic chemistry, but the large search space inherent in synthesis planning and the difficulty in predicting reaction conditions make it a significant challenge. We introduce a new Monte Carlo Tree Search (MCTS) variant that promotes balance between exploration and exploitation across the synthesis space. Together with a value network trained from reinforcement learning and a solvent-prediction neural network, our algorithm is comparable to the best MCTS variant (PUCT, similar to Google's Alpha Go) in finding valid synthesis pathways within a fixed searching time, and superior in identifying shorter routes with greener solvents under the same search conditions. In addition, with the same root compound visit count, our algorithm outperforms the PUCT MCTS by 16% in terms of determining successful routes. Overall the success rate is improved by 19.7% compared to the upper confidence bound applied to trees (UCT) MCTS method. Moreover, we improve 71.4% of the routes proposed by the PUCT MCTS variant in pathway length and choices of green solvents. The approach generally enables including Green Chemistry considerations in computer aided synthesis planning with potential applications in process development for fine chemicals or pharmaceuticals. |
| first_indexed | 2024-09-23T15:15:56Z |
| format | Article |
| id | mit-1721.1/135198 |
| institution | Massachusetts Institute of Technology |
| language | English |
| last_indexed | 2024-09-23T15:15:56Z |
| publishDate | 2021 |
| publisher | Royal Society of Chemistry (RSC) |
| record_format | dspace |
| spelling | mit-1721.1/1351982021-10-28T03:24:45Z Towards efficient discovery of green synthetic pathways with Monte Carlo tree search and reinforcement learning Wang, Xiaoxue Qian, Yujie Gao, Hanyu Coley, Connor W Mo, Yiming Barzilay, Regina Jensen, Klavs F © The Royal Society of Chemistry. Computer aided synthesis planning of synthetic pathways with green process conditions has become of increasing importance in organic chemistry, but the large search space inherent in synthesis planning and the difficulty in predicting reaction conditions make it a significant challenge. We introduce a new Monte Carlo Tree Search (MCTS) variant that promotes balance between exploration and exploitation across the synthesis space. Together with a value network trained from reinforcement learning and a solvent-prediction neural network, our algorithm is comparable to the best MCTS variant (PUCT, similar to Google's Alpha Go) in finding valid synthesis pathways within a fixed searching time, and superior in identifying shorter routes with greener solvents under the same search conditions. In addition, with the same root compound visit count, our algorithm outperforms the PUCT MCTS by 16% in terms of determining successful routes. Overall the success rate is improved by 19.7% compared to the upper confidence bound applied to trees (UCT) MCTS method. Moreover, we improve 71.4% of the routes proposed by the PUCT MCTS variant in pathway length and choices of green solvents. The approach generally enables including Green Chemistry considerations in computer aided synthesis planning with potential applications in process development for fine chemicals or pharmaceuticals. 2021-10-27T20:22:25Z 2021-10-27T20:22:25Z 2020 2020-12-01T18:03:00Z Article http://purl.org/eprint/type/JournalArticle https://hdl.handle.net/1721.1/135198 en 10.1039/d0sc04184j Chemical Science Creative Commons Attribution Noncommercial 3.0 unported license https://creativecommons.org/licenses/by-nc/3.0/ application/pdf Royal Society of Chemistry (RSC) Royal Society of Chemistry (RSC) |
| spellingShingle | Wang, Xiaoxue Qian, Yujie Gao, Hanyu Coley, Connor W Mo, Yiming Barzilay, Regina Jensen, Klavs F Towards efficient discovery of green synthetic pathways with Monte Carlo tree search and reinforcement learning |
| title | Towards efficient discovery of green synthetic pathways with Monte Carlo tree search and reinforcement learning |
| title_full | Towards efficient discovery of green synthetic pathways with Monte Carlo tree search and reinforcement learning |
| title_fullStr | Towards efficient discovery of green synthetic pathways with Monte Carlo tree search and reinforcement learning |
| title_full_unstemmed | Towards efficient discovery of green synthetic pathways with Monte Carlo tree search and reinforcement learning |
| title_short | Towards efficient discovery of green synthetic pathways with Monte Carlo tree search and reinforcement learning |
| title_sort | towards efficient discovery of green synthetic pathways with monte carlo tree search and reinforcement learning |
| url | https://hdl.handle.net/1721.1/135198 |
| work_keys_str_mv | AT wangxiaoxue towardsefficientdiscoveryofgreensyntheticpathwayswithmontecarlotreesearchandreinforcementlearning AT qianyujie towardsefficientdiscoveryofgreensyntheticpathwayswithmontecarlotreesearchandreinforcementlearning AT gaohanyu towardsefficientdiscoveryofgreensyntheticpathwayswithmontecarlotreesearchandreinforcementlearning AT coleyconnorw towardsefficientdiscoveryofgreensyntheticpathwayswithmontecarlotreesearchandreinforcementlearning AT moyiming towardsefficientdiscoveryofgreensyntheticpathwayswithmontecarlotreesearchandreinforcementlearning AT barzilayregina towardsefficientdiscoveryofgreensyntheticpathwayswithmontecarlotreesearchandreinforcementlearning AT jensenklavsf towardsefficientdiscoveryofgreensyntheticpathwayswithmontecarlotreesearchandreinforcementlearning |