Learning from Failure: Predicting Electronic Structure Calculation Outcomes with Machine Learning Models
High-throughput computational screening for chemical discovery mandates the automated and unsupervised simulation of thousands of new molecules and materials. In challenging materials spaces, such as open shell transition metal chemistry, characterization requires time-consuming first-principles sim...
| Main Authors: | , , , , |
|---|---|
| Other Authors: | |
| Format: | Article |
| Language: | English |
| Published: |
American Chemical Society (ACS)
2020
|
| Subjects: | |
| Online Access: | https://hdl.handle.net/1721.1/128282 |
| _version_ | 1826210416816029696 |
|---|---|
| author | Duan, Chenru Janet, Jon Paul Liu, Fang Nandy, Aditya Kulik, Heather Janine |
| author2 | Massachusetts Institute of Technology. Department of Chemical Engineering |
| author_facet | Massachusetts Institute of Technology. Department of Chemical Engineering Duan, Chenru Janet, Jon Paul Liu, Fang Nandy, Aditya Kulik, Heather Janine |
| author_sort | Duan, Chenru |
| collection | MIT |
| description | High-throughput computational screening for chemical discovery mandates the automated and unsupervised simulation of thousands of new molecules and materials. In challenging materials spaces, such as open shell transition metal chemistry, characterization requires time-consuming first-principles simulation that often necessitates human intervention. These calculations can frequently lead to a null result, e.g., the calculation does not converge or the molecule does not stay intact during a geometry optimization. To overcome this challenge toward realizing fully automated chemical discovery in transition metal chemistry, we have developed the first machine learning models that predict the likelihood of successful simulation outcomes. We train support vector machine and artificial neural network classifiers to predict simulation outcomes (i.e., geometry optimization result and degree of S 2 deviation) for a chosen electronic structure method based on chemical composition. For these static models, we achieve an area under the curve of at least 0.95, minimizing computational time spent on nonproductive simulations and therefore enabling efficient chemical space exploration. We introduce a metric of model uncertainty based on the distribution of points in the latent space to systematically improve model prediction confidence. In a complementary approach, we train a convolutional neural network classification model on simulation output electronic and geometric structure time series data. This dynamic model generalizes more readily than the static classifier by becoming more predictive as input simulation length increases. Finally, we describe approaches for using these models to enable autonomous job control in transition metal complex discovery. ©2019 American Chemical Society. |
| first_indexed | 2024-09-23T14:49:34Z |
| format | Article |
| id | mit-1721.1/128282 |
| institution | Massachusetts Institute of Technology |
| language | English |
| last_indexed | 2024-09-23T14:49:34Z |
| publishDate | 2020 |
| publisher | American Chemical Society (ACS) |
| record_format | dspace |
| spelling | mit-1721.1/1282822022-09-29T10:48:21Z Learning from Failure: Predicting Electronic Structure Calculation Outcomes with Machine Learning Models Duan, Chenru Janet, Jon Paul Liu, Fang Nandy, Aditya Kulik, Heather Janine Massachusetts Institute of Technology. Department of Chemical Engineering Massachusetts Institute of Technology. Department of Chemistry Physical and Theoretical Chemistry Computer Science Applications High-throughput computational screening for chemical discovery mandates the automated and unsupervised simulation of thousands of new molecules and materials. In challenging materials spaces, such as open shell transition metal chemistry, characterization requires time-consuming first-principles simulation that often necessitates human intervention. These calculations can frequently lead to a null result, e.g., the calculation does not converge or the molecule does not stay intact during a geometry optimization. To overcome this challenge toward realizing fully automated chemical discovery in transition metal chemistry, we have developed the first machine learning models that predict the likelihood of successful simulation outcomes. We train support vector machine and artificial neural network classifiers to predict simulation outcomes (i.e., geometry optimization result and degree of S 2 deviation) for a chosen electronic structure method based on chemical composition. For these static models, we achieve an area under the curve of at least 0.95, minimizing computational time spent on nonproductive simulations and therefore enabling efficient chemical space exploration. We introduce a metric of model uncertainty based on the distribution of points in the latent space to systematically improve model prediction confidence. In a complementary approach, we train a convolutional neural network classification model on simulation output electronic and geometric structure time series data. This dynamic model generalizes more readily than the static classifier by becoming more predictive as input simulation length increases. Finally, we describe approaches for using these models to enable autonomous job control in transition metal complex discovery. ©2019 American Chemical Society. DARPA grant (D18AP00039) Office of Naval Research grant (N00014-17-1-2956) Office of Naval Research grant (N00014-18-1-2434) National Science Foundation Major Research Instrumentation program (ACI-1429830) National Science Foundation grant (number ACI-1548562) 2020-11-02T15:47:52Z 2020-11-02T15:47:52Z 2019-03 2019-01 2019-08-22T16:09:12Z Article http://purl.org/eprint/type/JournalArticle 1549-9618 1549-9626 https://hdl.handle.net/1721.1/128282 Duan, Chenru et al., "Learning from Failure: Predicting Electronic Structure Calculation Outcomes with Machine Learning Models." Journal of Chemical Theory and Computation 15, 4 (April 2019): 2331–2345 doi. 10.1021/acs.jctc.9b00057 ©2019 Authors en https://dx.doi.org/10.1021/acs.jctc.9b00057 Journal of Chemical Theory and Computation Article is made available in accordance with the publisher's policy and may be subject to US copyright law. Please refer to the publisher's site for terms of use. application/pdf American Chemical Society (ACS) Other repository |
| spellingShingle | Physical and Theoretical Chemistry Computer Science Applications Duan, Chenru Janet, Jon Paul Liu, Fang Nandy, Aditya Kulik, Heather Janine Learning from Failure: Predicting Electronic Structure Calculation Outcomes with Machine Learning Models |
| title | Learning from Failure: Predicting Electronic Structure Calculation Outcomes with Machine Learning Models |
| title_full | Learning from Failure: Predicting Electronic Structure Calculation Outcomes with Machine Learning Models |
| title_fullStr | Learning from Failure: Predicting Electronic Structure Calculation Outcomes with Machine Learning Models |
| title_full_unstemmed | Learning from Failure: Predicting Electronic Structure Calculation Outcomes with Machine Learning Models |
| title_short | Learning from Failure: Predicting Electronic Structure Calculation Outcomes with Machine Learning Models |
| title_sort | learning from failure predicting electronic structure calculation outcomes with machine learning models |
| topic | Physical and Theoretical Chemistry Computer Science Applications |
| url | https://hdl.handle.net/1721.1/128282 |
| work_keys_str_mv | AT duanchenru learningfromfailurepredictingelectronicstructurecalculationoutcomeswithmachinelearningmodels AT janetjonpaul learningfromfailurepredictingelectronicstructurecalculationoutcomeswithmachinelearningmodels AT liufang learningfromfailurepredictingelectronicstructurecalculationoutcomeswithmachinelearningmodels AT nandyaditya learningfromfailurepredictingelectronicstructurecalculationoutcomeswithmachinelearningmodels AT kulikheatherjanine learningfromfailurepredictingelectronicstructurecalculationoutcomeswithmachinelearningmodels |