ACS Publications. Most Trusted. Most Cited. Most Read
My Activity

Figure 1Loading Img
RETURN TO ISSUEPREVMachine Learning and...Machine Learning and Deep LearningNEXT

Molecule Edit Graph Attention Network: Modeling Chemical Reactions as Sequences of Graph Edits

Cite this: J. Chem. Inf. Model. 2021, 61, 7, 3273–3284
Publication Date (Web):July 12, 2021
Copyright © 2021 American Chemical Society

    Article Views





    Read OnlinePDF (1 MB)
    Supporting Info (1)»


    Abstract Image

    The central challenge in automated synthesis planning is to be able to generate and predict outcomes of a diverse set of chemical reactions. In particular, in many cases, the most likely synthesis pathway cannot be applied due to additional constraints, which requires proposing alternative chemical reactions. With this in mind, we present Molecule Edit Graph Attention Network (MEGAN), an end-to-end encoder–decoder neural model. MEGAN is inspired by models that express a chemical reaction as a sequence of graph edits, akin to the arrow pushing formalism. We extend this model to retrosynthesis prediction (predicting substrates given the product of a chemical reaction) and scale it up to large data sets. We argue that representing the reaction as a sequence of edits enables MEGAN to efficiently explore the space of plausible chemical reactions, maintaining the flexibility of modeling the reaction in an end-to-end fashion and achieving state-of-the-art accuracy in standard benchmarks. Code and trained models are made available online at

    Supporting Information

    Jump To

    The Supporting Information is available free of charge at

    • Detailed description of the featurization used in the MEGAN model, list of graph actions used in experiments on the USPTO-50k data set, description of the algorithm used to break ties when determining the ground truth sequence of actions in training of the MEGAN model, and values of hyperparameters used in the experiments (PDF)

    Terms & Conditions

    Most electronic Supporting Information files are available without a subscription to ACS Web Editions. Such files may be downloaded by article for research use (if there is a public use license linked to the relevant article, that license may permit other uses). Permission may be obtained from ACS for other uses through requests via the RightsLink permission system:

    Cited By

    This article is cited by 23 publications.

    1. Zaiyun Lin, Shiqiu Yin, Lei Shi, Wenbiao Zhou, Yingsheng John Zhang. G2GT: Retrosynthesis Prediction with Graph-to-Graph Attention Neural Network and Self-Training. Journal of Chemical Information and Modeling 2023, 63 (7) , 1894-1905.
    2. Zhengkai Tu, Connor W. Coley. Permutation Invariant Graph-to-Sequence Model for Template-Free Retrosynthesis and Reaction Prediction. Journal of Chemical Information and Modeling 2022, 62 (15) , 3503-3513.
    3. Aditya Sonpal Mohammad Atif Faiz Afzal Yuling An Anand Chandrasekaran Mathew D. Halls . Benchmarking Machine Learning Descriptors for Crystals. , 111-126.
    4. Rocío Mercado, Esben J. Bjerrum, Ola Engkvist. Exploring Graph Traversal Algorithms in Graph-Based Molecular Generation. Journal of Chemical Information and Modeling 2022, 62 (9) , 2093-2100.
    5. Philipp Seidl, Philipp Renz, Natalia Dyubankova, Paulo Neves, Jonas Verhoeven, Jörg K. Wegner, Marwin Segler, Sepp Hochreiter, Günter Klambauer. Improving Few- and Zero-Shot Reaction Template Prediction Using Modern Hopfield Networks. Journal of Chemical Information and Modeling 2022, 62 (9) , 2111-2120.
    6. Esther Heid, Jiannan Liu, Andrea Aude, William H. Green. Influence of Template Size, Canonicalization, and Exclusivity for Retrosynthesis and Reaction Prediction Applications. Journal of Chemical Information and Modeling 2022, 62 (1) , 16-26.
    7. Shuan Chen, Yousung Jung. Deep Retrosynthetic Reaction Prediction using Local Reactivity and Global Attention. JACS Au 2021, 1 (10) , 1612-1620.
    8. Huihui Yang, Xiaosen Bai, Baolong Feng, Qinghua Wang, Li Meng, Fengzhong Wang, Yutang Wang. Application of Molecular Transformer approach for predicting the potential reactions to generate advanced glycation end products in infant formula. Food Chemistry 2023, 407 , 135143.
    9. Mikhail Andronov, Varvara Voinarovska, Natalia Andronova, Michael Wand, Djork-Arné Clevert, Jürgen Schmidhuber. Reagent prediction with a molecular transformer improves reaction data quality. Chemical Science 2023, 14 (12) , 3235-3246.
    10. Zhengkai Tu, Thijs Stuyver, Connor W. Coley. Predictive chemistry: machine learning for reaction deployment, reaction development, and reaction discovery. Chemical Science 2023, 14 (2) , 226-244.
    11. Xingyu Yan, Yin Zhang, Mingfang Huang, Xiaolian Yang, Yi Yan, Fang Hu. Graph-based medicine embedding learning via multiple attentions. Computers and Electrical Engineering 2023, 105 , 108494.
    12. Patrick Reiser, Marlen Neubert, André Eberhard, Luca Torresi, Chen Zhou, Chen Shao, Houssam Metni, Clint van Hoesel, Henrik Schopmans, Timo Sommer, Pascal Friederich. Graph neural networks for materials science and chemistry. Communications Materials 2022, 3 (1)
    13. Hua-Rui He, Jie Wang, Yunfei Liu, Feng Wu. Modeling Diverse Chemical Reactions for Single-step Retrosynthesis via Discrete Latent Variables. 2022, 717-726.
    14. Silvia Chines, Christiane Ehrt, Marco Potowski, Felix Biesenkamp, Lars Grützbach, Susanne Brunner, Frederik van den Broek, Shilpa Bali, Katja Ickstadt, Andreas Brunschweiger. Navigating chemical reaction space – application to DNA-encoded chemistry. Chemical Science 2022, 13 (37) , 11221-11231.
    15. Philippe Schwaller, Alain C. Vaucher, Ruben Laplaza, Charlotte Bunne, Andreas Krause, Clemence Corminboeuf, Teodoro Laino. Machine intelligence for chemical reaction space. WIREs Computational Molecular Science 2022, 12 (5)
    16. Shuan Chen, Yousung Jung. A generalized-template-based graph neural network for accurate organic reactivity prediction. Nature Machine Intelligence 2022, 4 (9) , 772-780.
    17. Chaochao Yan, Peilin Zhao, Chan Lu, Yang Yu, Junzhou Huang. RetroComposer: Composing Templates for Template-Based Retrosynthesis Prediction. Biomolecules 2022, 12 (9) , 1325.
    18. Zipeng Zhong, Jie Song, Zunlei Feng, Tiantao Liu, Lingxiang Jia, Shaolun Yao, Min Wu, Tingjun Hou, Mingli Song. Root-aligned SMILES: a tight representation for chemical reaction prediction. Chemical Science 2022, 13 (31) , 9023-9034.
    19. Samuel Genheden, Esben Bjerrum. PaRoutes: towards a framework for benchmarking retrosynthesis route predictions. Digital Discovery 2022, 1 (4) , 527-539.
    20. Yinjie Jiang, Yemin Yu, Ming Kong, Yu Mei, Luotian Yuan, Zhengxing Huang, Kun Kuang, Zhihua Wang, Huaxiu Yao, James Zou, Connor W. Coley, Ying Wei. Artificial Intelligence for Retrosynthesis Prediction. Engineering 2022, 39
    21. Sanggil Park, Herim Han, Hyungjun Kim, Sunghwan Choi. Machine Learning Applications for Chemical Reactions. Chemistry – An Asian Journal 2022, 17 (14)
    22. R.S.K. Vijayan, Jan Kihlberg, Jason B. Cross, Vasanthanathan Poongavanam. Enhancing preclinical drug discovery with artificial intelligence. Drug Discovery Today 2022, 27 (4) , 967-984.
    23. Nelson Vermeer, Andrew R. Brown. Generating Novel Furniture with Machine Learning. 2022, 401-416.

    Pair your accounts.

    Export articles to Mendeley

    Get article recommendations from ACS based on references in your Mendeley library.

    Pair your accounts.

    Export articles to Mendeley

    Get article recommendations from ACS based on references in your Mendeley library.

    You’ve supercharged your research process with ACS and Mendeley!

    STEP 1:
    Click to create an ACS ID

    Please note: If you switch to a different device, you may be asked to login again with only your ACS ID.

    Please note: If you switch to a different device, you may be asked to login again with only your ACS ID.

    Please note: If you switch to a different device, you may be asked to login again with only your ACS ID.

    Your Mendeley pairing has expired. Please reconnect