Predicting Organic Reaction Outcomes with Weisfeiler-Lehman Network

By Wengong Jin, Connor W. Coley, Regina Barzilay, Tommi S. Jaakkola. 2017

332 citations recorded by the source corpus.

The prediction of organic reaction outcomes is a fundamental problem in computational chemistry. Since a reaction may involve hundreds of atoms, fully exploring the space of possible transformations is intractable. The current solution utilizes reaction templates to limit the space, but it suffers from coverage and efficiency issues. In this paper, we propose 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. Since only a small number of atoms contribute to reaction center, we can directly enumerate candidate products. The generated candidates are scored by a Weisfeiler-Lehman Difference Network that models high-order interactions between changes occurring at nodes across the molecule. Our framework outperforms the top-performing template-based approach with a 10\% margin, while running orders of magnitude faster. Finally, we demonstrate that the model accuracy rivals the performance of domain experts.

Subfields: Artificial Intelligence, Computational Theory and Mathematics, Materials Chemistry

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Artificial Intelligence

In Computational Theory and Mathematics

In Materials Chemistry

Citations in the corpus

Hierarchical Graph Representation Learning with Differentiable Pooling (2018, methodology): “GNNs have been applied to a wide variety of tasks, including node classification [16, 21], link prediction [31], graph classification [7, 11, 39], and chemoinformatics [14, 19, 27, 28, 32].”
Learning Multimodal Graph-to-Graph Translation for Molecular Optimization (2018, methodology): “Graph encoders have been applied to social network analysis (Kipf & Welling, 2016; Hamilton et al., 2017) and chemistry (Kearnes et al., 2016; Gilmer et al., 2017; Sch ¨ utt et al., 2017; Jin et al., 2017).”
Junction Tree Variational Autoencoder for Molecular Graph Generation (2018, methodology): “For applications, graph neural networks are used in semi-supervised classiﬁcation (Kipf & Welling, 2016), computer vision (Monti et al., 2016), and chemical domains (Kearnes et al., 2016; Sch ¨ utt et al., 2017; Jin et al., 2017).”
TorchDrug: A Powerful and Flexible Machine Learning Platform for Drug Discovery (2022, background): “, 2020b) , reaction prediction (Jin et al., 2017; Bradshaw et al., 2018), retrosynthesis prediction (Dai et al.”
A robotic platform for flow synthesis of organic compounds informed by AI planning (2019, methodology): “Pathways that consist of reaction steps considered plausible by the binary classifier can be evaluated by themore rigorous forward prediction model, which explicitly generates product molecules based on the provided reactants and reaction conditions (37, 38); the exact model used is described in det”
Using Machine Learning To Predict Suitable Conditions for Organic Reactions (2018, background): “20,21 More importantly, reaction conditions are necessary to evaluate opportunities for one-pot synthesis, telescoping in ﬂ ow, and amenability to the use of “ green ” solvents for sustainability.”
Chemformer: a pre-trained transformer for computational chemistry (2021, unknown): “For the direct synthesis prediction task we made use of the benchmark USPTO-MIT dataset [33], which contains approximately 470 000 reactions originally extracted from patents [34].”
Artificial intelligence in drug discovery: recent advances and future perspectives (2021, methodology): “by viewing the chemical reaction prediction problem as a graph transformation task [181,182].”
Molecular Transformer: A Model for Uncertainty-Calibrated Chemical Reaction Prediction (2018, methodology): “The USPTO_MIT dataset was ﬁltered and split by Jin et al [28].”
Graph convolutional networks for computational drug development and discovery (2019, unknown): “In [43], Jin et al . proposed a template-free approach for reaction prediction leveraging graph neural networks.”
GNNExplainer: Generating Explanations for Graph Neural Networks (2019, methodology): “14, 46] or graph classification [12, 16, 46]. These tasks occur in domains where the graph structure is ubiquitous, such as social networks [5], content graphs [45], biology [3], and chemoinformatics [16, 25, 54]. Recent instances of GNN models have been proposed to augment the interpretability prop”
Therapeutics Data Commons: Machine Learning Datasets and Tasks for Drug Discovery and Development (2021, background): “Reaction outcome prediction is to predict the reaction products given a set of reactants (Jin et al. 2017).”
“Found in Translation”: predicting outcomes of complex organic chemistry reactions using neural sequence-to-sequence models (2017, methodology): “Step Example (entry 23738, Jin’s USPTO test set [17]) reactants > reagents > products”
Enabling late-stage drug diversification by high-throughput experimentation with geometric deep learning (2023, background): “…medicinal chemistry include Article https://doi.org/10.1038/s41557-023-01360-5 various machine learning methods developed for chemical reaction planning 23,29,30 , GNNs have been successfully employed for retrosynthe-sis planning, regioselectivity prediction and reaction product prediction 31–34 .”
Geometric deep learning on molecular representations (2021, unknown): “For modelling tasks that are usually characterized by the need for engineered rules (for example, molecule construction for de novo design and reactive site specification for CASP), GDL extends the existing methodological repertoire.”
Constrained Bayesian Optimization for Automatic Chemical Design (2017, methodology): “This text-based format facilitates the use of tools from natural language processing for applications such as organic chemistry reaction prediction Schwaller et al. [2017], Jin et al. [2017].”