Molecule Edit Graph Attention Network: Modeling Chemical Reactions as Sequences of Graph Edits (2020-06-27T00:00:00.000000Z)

TL;DR

It is argued 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.

Abstract

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

Authors

Stanislaw Jastrzebski

5 papers

Mikolaj Sacha

1 papers

Mikolaj Blaz

1 papers

TL;DR

Abstract

Authors

References82 items

Protein design and variant prediction using autoregressive generative models

Expanding functional protein sequence spaces using generative adversarial networks

RegioSQM20: improved prediction of the regioselectivity of electrophilic aromatic substitutions

Data augmentation and transfer learning strategies for reaction prediction in low chemical data regimes

Decoding and Diversity in Machine Translation

RetroGNN: Approximating Retrosynthesis by Graph Neural Networks for De Novo Drug Design

RetroXpert: Decompose Retrosynthesis Prediction Like A Chemist

Retrosynthetic accessibility score (RAscore) – rapid machine learned synthesizability classification from AI driven retrosynthetic planning†

Transfer learning enables the molecular transformer to predict regio- and stereoselective reactions on carbohydrates

RetroPrime: A Chemistry-Inspired and Transformer-based Method for Retrosynthesis Predictions

De Novo Protein Design for Novel Folds Using Guided Conditional Wasserstein Generative Adversarial Networks

Regio-selectivity prediction with a machine-learned reaction representation and on-the-fly quantum mechanical descriptors

RetroXpert: Decompose Retrosynthesis Prediction like a Chemist

Learning Graph Models for Template-Free Retrosynthesis

Generative Network Complex for the Automated Generation of Drug-like Molecules

Learning To Navigate The Synthetically Accessible Chemical Space Using Reinforcement Learning

Current and Future Roles of Artificial Intelligence in Medicinal Chemistry Synthesis

A Graph to Graphs Framework for Retrosynthesis Prediction

State-of-the-art augmented NLP transformer models for direct and single-step retrosynthesis

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

Automatic retrosynthetic route planning using template-free models

The Synthesizability of Molecules Proposed by Generative Models

Data Augmentation and Pretraining for Template-Based Retrosynthetic Prediction in Computer-Aided Synthesis Planning

Retrosynthesis Prediction with Conditional Graph Logic Network

Predicting Retrosynthetic Reactions Using Self-Corrected Transformer Neural Networks

ProGAN: Protein solubility generative adversarial nets for data augmentation in DNN framework

Molecular Transformer unifies reaction prediction and retrosynthesis across pharma chemical space.

Learning to Make Generalizable and Diverse Predictions for Retrosynthesis

The logic of translating chemical knowledge into machine-processable forms: a modern playground for physical-organic chemistry

Prediction and Interpretable Visualization of Retrosynthetic Reactions Using Graph Convolutional Networks

RamaNet: Computational de novo helical protein backbone design using a long short-term memory generative neural network

The digitization of organic synthesis

A Transformer Model for Retrosynthesis

Insertion-based Decoding with Automatically Inferred Generation Order

Enhancing Retrosynthetic Reaction Prediction with Deep Learning Using Multiscale Reaction Classification

Graph Transformation Policy Network for Chemical Reaction Prediction

Molecular Transformer: A Model for Uncertainty-Calibrated Chemical Reaction Prediction

Middle-Out Decoding

A graph-convolutional neural network model for the prediction of chemical reactivity

Sequence to Sequence Mixture Model for Diverse Machine Translation

Why are Sequence-to-Sequence Models So Dull? Understanding the Low-Diversity Problem of Chatbots

Graph Convolutional Policy Network for Goal-Directed Molecular Graph Generation

A Generative Model For Electron Paths

Machine Learning in Computer-Aided Synthesis Planning.

Organic synthesis provides opportunities to transform drug discovery

Chematica: A Story of Computer Code That Started to Think like a Chemist

GraphRNN: Generating Realistic Graphs with Deep Auto-regressive Models

Generating Focused Molecule Libraries for Drug Discovery with Recurrent Neural Networks

Deep reinforcement learning for de novo drug design

Computer-Assisted Retrosynthesis Based on Molecular Similarity

Graph Attention Networks

Crystal Graph Convolutional Neural Networks for an Accurate and Interpretable Prediction of Material Properties.

Learning Graph-Level Representation for Drug Discovery

Predicting Organic Reaction Outcomes with Weisfeiler-Lehman Network

Planning chemical syntheses with deep neural networks and symbolic AI

Attention is All you Need

Retrosynthetic Reaction Prediction Using Neural Sequence-to-Sequence Models

Neural-Symbolic Machine Learning for Retrosynthesis and Reaction Prediction.

Molecular de-novo design through deep reinforcement learning

Prediction of Organic Reaction Outcomes Using Machine Learning

What's What: The (Nearly) Definitive Guide to Reaction Role Assignment

Synergies Between Quantum Mechanics and Machine Learning in Reaction Prediction

Neural Networks for the Prediction of Organic Chemistry Reactions

Automatic Chemical Design Using a Data-Driven Continuous Representation of Molecules

Modelling Chemical Reasoning to Predict Reactions

Convolutional Networks on Graphs for Learning Molecular Fingerprints

Adam: A Method for Stochastic Optimization

Automated discovery of chemically reasonable elementary reaction steps

Sequence Transduction with Recurrent Neural Networks

Extraction of chemical structures and reactions from the literature

Route Designer: A Retrosynthetic Analysis Tool Utilizing Automated Retrosynthetic Rule Generation

Found in translation

Computer-aided organic synthesis.

SMILES, a chemical language and information system. 1. Introduction to methodology and encoding rules

Computer-assisted mechanistic evaluation of organic reactions. 1. Overview

Computer-assisted design of complex organic syntheses.