Autonomous Design of New Chemical Reactions Using a Variational Autoencoder

Overview

Journal Commun Chem

Publisher Springer Nature

Specialty Chemistry

Date 2023 Jan 25

PMID 36697652

Authors

Robert Tempke

Terence Musho

Affiliations

Soon will be listed here.

Abstract

Artificial intelligence based chemistry models are a promising method of exploring chemical reaction design spaces. However, training datasets based on experimental synthesis are typically reported only for the optimal synthesis reactions. This leads to an inherited bias in the model predictions. Therefore, robust datasets that span the entirety of the solution space are necessary to remove inherited bias and permit complete training of the space. In this study, an artificial intelligence model based on a Variational AutoEncoder (VAE) has been developed and investigated to synthetically generate continuous datasets. The approach involves sampling the latent space to generate new chemical reactions. This developed technique is demonstrated by generating over 7,000,000 new reactions from a training dataset containing only 7,000 reactions. The generated reactions include molecular species that are larger and more diverse than the training set.

Citing Articles

Fuzz Testing Molecular Representation Using Deep Variational Anomaly Generation.

Nogueira V, Sharma R, Guido R, Keiser M J Chem Inf Model. 2025; 65(4):1911-1927.

PMID: 39908426 PMC: 11863373. DOI: 10.1021/acs.jcim.4c01876.

Into the Unknown: How Computation Can Help Explore Uncharted Material Space.

Mroz A, Posligua V, Tarzia A, Wolpert E, Jelfs K J Am Chem Soc. 2022; 144(41):18730-18743.

PMID: 36206484 PMC: 9585593. DOI: 10.1021/jacs.2c06833.

References

Gajendran S, D M, Sugumaran V . Character level and word level embedding with bidirectional LSTM - Dynamic recurrent neural network for biomedical named entity recognition from literature. J Biomed Inform. 2020; 112:103609. DOI: 10.1016/j.jbi.2020.103609. View

Kim S, Chen J, Cheng T, Gindulyte A, He J, He S . PubChem in 2021: new data content and improved web interfaces. Nucleic Acids Res. 2020; 49(D1):D1388-D1395. PMC: 7778930. DOI: 10.1093/nar/gkaa971. View

Jia X, Lynch A, Huang Y, Danielson M, Langat I, Milder A . Anthropogenic biases in chemical reaction data hinder exploratory inorganic synthesis. Nature. 2019; 573(7773):251-255. DOI: 10.1038/s41586-019-1540-5. View

Karwath A, De Raedt L . SMIREP: predicting chemical activity from SMILES. J Chem Inf Model. 2006; 46(6):2432-44. DOI: 10.1021/ci060159g. View

Kovacs D, McCorkindale W, Lee A . Quantitative interpretation explains machine learning models for chemical reaction prediction and uncovers bias. Nat Commun. 2021; 12(1):1695. PMC: 7966799. DOI: 10.1038/s41467-021-21895-w. View

Stewart J . Optimization of parameters for semiempirical methods V: modification of NDDO approximations and application to 70 elements. J Mol Model. 2007; 13(12):1173-213. PMC: 2039871. DOI: 10.1007/s00894-007-0233-4. View

Gale E, Durand D . Improving reaction prediction. Nat Chem. 2020; 12(6):509-510. DOI: 10.1038/s41557-020-0478-4. View

Pinheiro G, Mucelini J, Soares M, Prati R, Da Silva J, Quiles M . Machine Learning Prediction of Nine Molecular Properties Based on the SMILES Representation of the QM9 Quantum-Chemistry Dataset. J Phys Chem A. 2020; 124(47):9854-9866. DOI: 10.1021/acs.jpca.0c05969. View

Glavatskikh M, Leguy J, Hunault G, Cauchy T, Da Mota B . Dataset's chemical diversity limits the generalizability of machine learning predictions. J Cheminform. 2021; 11(1):69. PMC: 6852905. DOI: 10.1186/s13321-019-0391-2. View

10.

Carroll H, Toumpakari Z, Johnson L, Betts J . The perceived feasibility of methods to reduce publication bias. PLoS One. 2017; 12(10):e0186472. PMC: 5655535. DOI: 10.1371/journal.pone.0186472. View

11.

Gao W, Coley C . The Synthesizability of Molecules Proposed by Generative Models. J Chem Inf Model. 2020; 60(12):5714-5723. DOI: 10.1021/acs.jcim.0c00174. View

12.

Shields B, Stevens J, Li J, Parasram M, Damani F, Martinez Alvarado J . Bayesian reaction optimization as a tool for chemical synthesis. Nature. 2021; 590(7844):89-96. DOI: 10.1038/s41586-021-03213-y. View

13.

Cova T, Pais A . Deep Learning for Deep Chemistry: Optimizing the Prediction of Chemical Patterns. Front Chem. 2020; 7:809. PMC: 6988795. DOI: 10.3389/fchem.2019.00809. View

14.

Mater A, Coote M . Deep Learning in Chemistry. J Chem Inf Model. 2019; 59(6):2545-2559. DOI: 10.1021/acs.jcim.9b00266. View

15.

Kayala M, Baldi P . ReactionPredictor: prediction of complex chemical reactions at the mechanistic level using machine learning. J Chem Inf Model. 2012; 52(10):2526-40. DOI: 10.1021/ci3003039. View

16.

Schwaller P, Petraglia R, Zullo V, Nair V, Haeuselmann R, Pisoni R . Predicting retrosynthetic pathways using transformer-based models and a hyper-graph exploration strategy. Chem Sci. 2021; 11(12):3316-3325. PMC: 8152799. DOI: 10.1039/c9sc05704h. View

17.

Jinich A, Sanchez-Lengeling B, Ren H, Harman R, Aspuru-Guzik A . A Mixed Quantum Chemistry/Machine Learning Approach for the Fast and Accurate Prediction of Biochemical Redox Potentials and Its Large-Scale Application to 315 000 Redox Reactions. ACS Cent Sci. 2019; 5(7):1199-1210. PMC: 6661861. DOI: 10.1021/acscentsci.9b00297. View

18.

Segler M, Preuss M, Waller M . Planning chemical syntheses with deep neural networks and symbolic AI. Nature. 2018; 555(7698):604-610. DOI: 10.1038/nature25978. View

19.

Struble T, Alvarez J, Brown S, Chytil M, Cisar J, DesJarlais R . Current and Future Roles of Artificial Intelligence in Medicinal Chemistry Synthesis. J Med Chem. 2020; 63(16):8667-8682. PMC: 7457232. DOI: 10.1021/acs.jmedchem.9b02120. View

20.

Schwaller P, Laino T, Gaudin T, Bolgar P, Hunter C, Bekas C . Molecular Transformer: A Model for Uncertainty-Calibrated Chemical Reaction Prediction. ACS Cent Sci. 2019; 5(9):1572-1583. PMC: 6764164. DOI: 10.1021/acscentsci.9b00576. View