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

  title={Molecule Edit Graph Attention Network: Modeling Chemical Reactions as Sequences of Graph Edits},
  author={Mikolaj Sacha and Mikolaj Blaz and Piotr Byrski and Pawel Wlodarczyk-Pruszynski and Stanislaw Jastrzebski},
  journal={Journal of chemical information and modeling},
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… 

Permutation invariant graph-to-sequence model for template-free retrosynthesis and reaction prediction

A novel Graph2SMILES model that combines the power of Transformer models for text generation with the permutation invariance of molecular graph encoders that mitigates the need for input data augmentation is described.

Modern Hopfield Networks for Few- and Zero-Shot Reaction Template Prediction

This study model single-step retrosynthesis in a template-based approach using modern Hopfield networks (MHNs) to associate different modalities, reaction templates and molecules, which allows the model to leverage structural information about reaction templates.

Deep Retrosynthetic Reaction Prediction using Local Reactivity and Global Attention

A local retrosynthesis framework called LocalRetro is proposed, motivated by the chemical intuition that the molecular changes occur mostly locally during the chemical reactions, and shows a promising 89.5 and 99.2% round-trip accuracy at top-1 and top-5 predictions for the USPTO-50K dataset containing 50 016 reactions.

Non-autoregressive electron flow generation for reaction prediction

A novel decoder is devised that avoids such sequential generating and predicts the reaction in a Non-Autoregressive manner, with state-of-the-art top-1 accuracy and comparable performance on Top-K sampling.

Learning to Extend Molecular Scaffolds with Structural Motifs

This work proposes MoLeR, a graph-based model that naturally supports scaffolds as initial seed of the generative procedure, which is possible because it is not conditioned on the generation history, and performs comparably to state-of-the-art methods on unconstrained molecular optimization tasks, and outperforms them on scaffold-based tasks.

Non-Autoregressive Electron Redistribution Modeling for Reaction Prediction

This work formulate a reaction as an arbitrary electron flow and predict it with a novel multipointer decoding network and achieves at least 27 times inference speedup over the state-of-the-art methods.

Quantitative interpretation explains machine learning models for chemical reaction prediction and uncovers bias

This paper quantitatively interpret the Molecular Transformer, the state-of-the-art model for reaction prediction, and develops a framework to attribute predicted reaction outcomes both to specific parts of reactants, and to reactions in the training set.

Automation and computer-assisted planning for chemical synthesis

This Primer summarizes the most relevant aspects of chemical synthesis in this information age for those looking to understand the software, hardware and data and how these are used to enable retrosynthetic logic, reaction prediction and automation.

Exploring Graph Traversal Algorithms in Graph-Based Molecular Generation

Here, we explore the impact of different graph traversal algorithms on molecular graph generation. We do this by training a graph-based deep molecular generative model to build structures using a

Influence of Template Size, Canonicalization, and Exclusivity for Retrosynthesis and Reaction Prediction Applications

A general and computationally efficient framework to deduplicate and hierarchically correct templates is developed and performance improved considerably for both heuristic and machine learning template ranking models, as well as multistep retrosynthetic planning models.



Graph Transformation Policy Network for Chemical Reaction Prediction

This work proposes Graph Transformation Policy Network (GTPN), a novel generic method that combines the strengths of graph neural networks and reinforcement learning to learn reactions directly from data with minimal chemical knowledge.

Learning Graph Models for Template-Free Retrosynthesis

A graph-based approach that capitalizes on the idea that the graph topology of precursor molecules is largely unaltered during the reaction, greatly simplifying the underlying architecture and improving its ability to generalize.

A Generative Model For Electron Paths

A method to extract approximate reaction paths from any dataset of atom-mapped reaction SMILES strings and shows that the model achieves excellent performance on an important subset of the USPTO reaction dataset, comparing favorably to the strongest baselines.

Retrosynthesis Prediction with Conditional Graph Logic Network

The Conditional Graph Logic Network is proposed, a conditional graphical model built upon graph neural networks that learns when rules from reaction templates should be applied, implicitly considering whether the resulting reaction would be both chemically feasible and strategic.

A Graph to Graphs Framework for Retrosynthesis Prediction

Experimental results show that G2Gs significantly outperforms existing template-free approaches by up to 63% in terms of the top-1 accuracy and achieves a performance close to that of state-of-the-art template based approaches, but does not require domain knowledge and is much more scalable.

Enhancing Retrosynthetic Reaction Prediction with Deep Learning Using Multiscale Reaction Classification

The multiscale approach was unable to outperform state-of-the-art rule-based systems curated by expert chemists, but represents a marked enhancement in retrosynthetic analysis and can be easily adapted for use in a range of artificial intelligence strategies.

Predicting retrosynthetic pathways using transformer-based models and a hyper-graph exploration strategy†

An extension of the Molecular Transformer model combined with a hyper-graph exploration strategy for automatic retrosynthesis route planning without human intervention is presented and the end-to-end framework has an excellent performance with few weaknesses related to the training data.

Modelling Chemical Reasoning to Predict Reactions

A model that mimics chemical reasoning, and formalises reaction prediction as finding missing links in a knowledge graph that outperforms a rule-based expert system in the reaction prediction task for 180 000 randomly selected binary reactions.

Graph Convolutional Policy Network for Goal-Directed Molecular Graph Generation

Graph Convolutional Policy Network (GCPN) is proposed, a general graph convolutional network based model for goal-directed graph generation through reinforcement learning that can achieve 61% improvement on chemical property optimization over state-of-the-art baselines while resembling known molecules, and achieve 184% improved on the constrained property optimization task.

Predicting Organic Reaction Outcomes with Weisfeiler-Lehman Network

This paper proposes a template-free approach to efficiently explore the space of product molecules by first pinpointing the reaction center -- the set of nodes and edges where graph edits occur, and demonstrates that the model accuracy rivals the performance of domain experts.