A graph-convolutional neural network model for the prediction of chemical reactivity
© 2019 The Royal Society of Chemistry. We present a supervised learning approach to predict the products of organic reactions given their reactants, reagents, and solvent(s). The prediction task is factored into two stages comparable to manual expert approaches: considering possible sites of reactiv...
| Main Authors: | , , , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Royal Society of Chemistry (RSC)
2021
|
| Online Access: | https://hdl.handle.net/1721.1/135165 |
| _version_ | 1826188124897673216 |
|---|---|
| author | Coley, Connor W Jin, Wengong Rogers, Luke Jamison, Timothy F Jaakkola, Tommi S Green, William H Barzilay, Regina Jensen, Klavs F |
| author_facet | Coley, Connor W Jin, Wengong Rogers, Luke Jamison, Timothy F Jaakkola, Tommi S Green, William H Barzilay, Regina Jensen, Klavs F |
| author_sort | Coley, Connor W |
| collection | MIT |
| description | © 2019 The Royal Society of Chemistry. We present a supervised learning approach to predict the products of organic reactions given their reactants, reagents, and solvent(s). The prediction task is factored into two stages comparable to manual expert approaches: considering possible sites of reactivity and evaluating their relative likelihoods. By training on hundreds of thousands of reaction precedents covering a broad range of reaction types from the patent literature, the neural model makes informed predictions of chemical reactivity. The model predicts the major product correctly over 85% of the time requiring around 100 ms per example, a significantly higher accuracy than achieved by previous machine learning approaches, and performs on par with expert chemists with years of formal training. We gain additional insight into predictions via the design of the neural model, revealing an understanding of chemistry qualitatively consistent with manual approaches. |
| first_indexed | 2024-09-23T07:54:58Z |
| format | Article |
| id | mit-1721.1/135165 |
| institution | Massachusetts Institute of Technology |
| language | English |
| last_indexed | 2024-09-23T07:54:58Z |
| publishDate | 2021 |
| publisher | Royal Society of Chemistry (RSC) |
| record_format | dspace |
| spelling | mit-1721.1/1351652021-10-28T05:07:10Z A graph-convolutional neural network model for the prediction of chemical reactivity Coley, Connor W Jin, Wengong Rogers, Luke Jamison, Timothy F Jaakkola, Tommi S Green, William H Barzilay, Regina Jensen, Klavs F © 2019 The Royal Society of Chemistry. We present a supervised learning approach to predict the products of organic reactions given their reactants, reagents, and solvent(s). The prediction task is factored into two stages comparable to manual expert approaches: considering possible sites of reactivity and evaluating their relative likelihoods. By training on hundreds of thousands of reaction precedents covering a broad range of reaction types from the patent literature, the neural model makes informed predictions of chemical reactivity. The model predicts the major product correctly over 85% of the time requiring around 100 ms per example, a significantly higher accuracy than achieved by previous machine learning approaches, and performs on par with expert chemists with years of formal training. We gain additional insight into predictions via the design of the neural model, revealing an understanding of chemistry qualitatively consistent with manual approaches. 2021-10-27T20:11:03Z 2021-10-27T20:11:03Z 2019 2019-05-07T18:39:30Z Article http://purl.org/eprint/type/JournalArticle https://hdl.handle.net/1721.1/135165 en 10.1039/c8sc04228d Chemical Science Creative Commons Attribution 3.0 unported license https://creativecommons.org/licenses/by/3.0/ application/pdf Royal Society of Chemistry (RSC) Royal Society of Chemistry (RSC) |
| spellingShingle | Coley, Connor W Jin, Wengong Rogers, Luke Jamison, Timothy F Jaakkola, Tommi S Green, William H Barzilay, Regina Jensen, Klavs F A graph-convolutional neural network model for the prediction of chemical reactivity |
| title | A graph-convolutional neural network model for the prediction of chemical reactivity |
| title_full | A graph-convolutional neural network model for the prediction of chemical reactivity |
| title_fullStr | A graph-convolutional neural network model for the prediction of chemical reactivity |
| title_full_unstemmed | A graph-convolutional neural network model for the prediction of chemical reactivity |
| title_short | A graph-convolutional neural network model for the prediction of chemical reactivity |
| title_sort | graph convolutional neural network model for the prediction of chemical reactivity |
| url | https://hdl.handle.net/1721.1/135165 |
| work_keys_str_mv | AT coleyconnorw agraphconvolutionalneuralnetworkmodelforthepredictionofchemicalreactivity AT jinwengong agraphconvolutionalneuralnetworkmodelforthepredictionofchemicalreactivity AT rogersluke agraphconvolutionalneuralnetworkmodelforthepredictionofchemicalreactivity AT jamisontimothyf agraphconvolutionalneuralnetworkmodelforthepredictionofchemicalreactivity AT jaakkolatommis agraphconvolutionalneuralnetworkmodelforthepredictionofchemicalreactivity AT greenwilliamh agraphconvolutionalneuralnetworkmodelforthepredictionofchemicalreactivity AT barzilayregina agraphconvolutionalneuralnetworkmodelforthepredictionofchemicalreactivity AT jensenklavsf agraphconvolutionalneuralnetworkmodelforthepredictionofchemicalreactivity AT coleyconnorw graphconvolutionalneuralnetworkmodelforthepredictionofchemicalreactivity AT jinwengong graphconvolutionalneuralnetworkmodelforthepredictionofchemicalreactivity AT rogersluke graphconvolutionalneuralnetworkmodelforthepredictionofchemicalreactivity AT jamisontimothyf graphconvolutionalneuralnetworkmodelforthepredictionofchemicalreactivity AT jaakkolatommis graphconvolutionalneuralnetworkmodelforthepredictionofchemicalreactivity AT greenwilliamh graphconvolutionalneuralnetworkmodelforthepredictionofchemicalreactivity AT barzilayregina graphconvolutionalneuralnetworkmodelforthepredictionofchemicalreactivity AT jensenklavsf graphconvolutionalneuralnetworkmodelforthepredictionofchemicalreactivity |