mlsanityData to 7 Oct 2026

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

arXiv 2006.15426 ↗First posted 27 Jun 2020cs.LG physics.chem-ph stat.MLcited by 87

Not on the map. Why?

Abstract (arXiv)

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 datasets. 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 https://github.com/molecule-one/megan.

Key facts. Cited by 87 papers in the map (data through 7 Oct 2026). First posted on arXiv 27 Jun 2020.

Builds on

No reference list yet.

Built on by

Cited by year (2026 to 7 Oct)

Citing papers per year: 2021: 7; 2022: 14; 2023: 16; 2024: 14; 2025: 20; 2026: 16.

87 papers.

Sort:

By area:

  1. 2021Therapeutics Data Commons: Machine Learning Datasets and Tasks for Drug Discovery and Developmentcited by 258
  2. 2022Graph neural networks for materials science and chemistrycited by 181
  3. 2021Learning to Extend Molecular Scaffolds with Structural Motifscited by 70
  4. 2021Permutation invariant graph-to-sequence model for template-free retrosynthesis and reaction predictioncited by 66
  5. 2022Root-aligned SMILES: A Tight Representation for Chemical Reaction Predictioncited by 49
  6. 2023RetroBridge: Modeling Retrosynthesis with Markov Bridgescited by 35
  7. 2022Towards Data-and Knowledge-Driven Artificial Intelligence: A Survey on Neuro-Symbolic Computingcited by 34
  8. 2025KL-Regularized Reinforcement Learning is Designed to Mode Collapsecited by 29
  9. 2023Re-evaluating Retrosynthesis Algorithms with Syntheseuscited by 26
  10. 2022Bidirectional Generation of Structure and Properties Through a Single Molecular Foundation Modelcited by 25
  11. 2023ReLM: Leveraging Language Models for Enhanced Chemical Reaction Predictioncited by 24
  12. 2024Double-Ended Synthesis Planning with Goal-Constrained Bidirectional Searchcited by 24
38 more
  1. 2022Graph Neural Networks for Moleculescited by 21
  2. 2024ReactXT: Understanding Molecular "Reaction-ship" via Reaction-Contextualized Molecule-Text Pretrainingcited by 19
  3. 2022Proactive Robot Assistance via Spatio-Temporal Object Modelingcited by 18
  4. 2021Non-Autoregressive Electron Redistribution Modeling for Reaction Predictioncited by 17
  5. 2023Retro-fallback: retrosynthetic planning in an uncertain worldcited by 16
  6. 2023Retrosynthesis Prediction with Local Template Retrievalcited by 15
  7. 2022FusionRetro: Molecule Representation Fusion via In-Context Learning for Retrosynthetic Planningcited by 14
  8. 2021Modern Hopfield Networks for Few- and Zero-Shot Reaction Template Predictioncited by 13
  9. 2021RetroComposer: Composing Templates for Template-Based Retrosynthesis Predictioncited by 13
  10. 2022SemiRetro: Semi-template framework boosts deep retrosynthesis predictioncited by 13
  11. 2025LLM-Augmented Chemical Synthesis and Design Decision Programscited by 13
  12. 2025Graph Neural Networks in Modern AI-aided Drug Discoverycited by 12
  13. 2024DirectMultiStep: Direct Route Generation for Multistep Retrosynthesiscited by 11
  14. 2023Deep Learning Methods for Small Molecule Drug Discovery: A Surveycited by 10
  15. 2024RetroGFN: Diverse and Feasible Retrosynthesis using GFlowNetscited by 10
  16. 2023Node-Aligned Graph-to-Graph (NAG2G): Elevating Template-Free Deep Learning Approaches in Single-Step Retrosynthesiscited by 9
  17. 2022Retroformer: Pushing the Limits of Interpretable End-to-end Retrosynthesis Transformercited by 8
  18. 2022MARS: A Motif-based Autoregressive Model for Retrosynthesis Predictioncited by 8
  19. 2022Mind the Retrosynthesis Gap: Bridging the divide between Single-step and Multi-step Retrosynthesis Predictioncited by 7
  20. 2023RetroDiff: Retrosynthesis as Multi-stage Distribution Interpolationcited by 7
  21. 2024Equivariant Denoisers Cannot Copy Graphs: Align Your Graph Diffusion Modelscited by 7
  22. 2024Tango*: Constrained synthesis planning using chemically informed value functionscited by 7
  23. 2023Doubly Stochastic Graph-based Non-autoregressive Reaction Predictioncited by 6
  24. 2024UAlign: Pushing the Limit of Template-free Retrosynthesis Prediction with Unsupervised SMILES Alignmentcited by 6
  25. 2025Challenging reaction prediction models to generalize to novel chemistrycited by 6
  26. 2021Unassisted Noise Reduction of Chemical Reaction Data Setscited by 5
  27. 2023Recent advances in artificial intelligence for retrosynthesiscited by 5
  28. 2024Evaluating Molecule Synthesizability via Retrosynthetic Planning and Reaction Predictioncited by 5
  29. 2024Chemist-aligned retrosynthesis by ensembling diverse inductive bias modelscited by 5
  30. 2022mathsfG2Retro as a Two-Step Graph Generative Models for Retrosynthesis Predictioncited by 4
  31. 2023A Unified View of Deep Learning for Reaction and Retrosynthesis Prediction: Current Status and Future Challengescited by 4
  32. 2024Accelerating the inference of string generation-based chemical reaction models for industrial applicationscited by 4
  33. 2022Modeling Diverse Chemical Reactions for Single-step Retrosynthesis via Discrete Latent Variablescited by 3
  34. 2024T-Rex: Text-assisted Retrosynthesis Predictioncited by 3
  35. 2025RetroDFM-R: Reasoning-Driven Retrosynthesis Prediction with Large Language Models via Reinforcement Learningcited by 3
  36. 2025Retro-Expert: Collaborative Reasoning for Interpretable Retrosynthesiscited by 3
  37. 2025Trustworthy Retrosynthesis: Eliminating Hallucinations with a Diverse Ensemble of Reaction Scorerscited by 3
  38. 2026Order Matters in Retrosynthesis: Structure-aware Generation via Reaction-Center-Guided Discrete Flow Matchingcited by 3

50 most cited of 87 All 87 (JSON)