Abstract:Chemical reaction prediction, encompassing forward synthesis and retrosynthesis, stands as a fundamental challenge in organic synthesis. A widely adopted computational approach frames synthesis prediction as a sequence-to-sequence translation task, using the common SMILES representation for molecules. Current evaluation of machine learning methods for retrosynthesis assume perfect training data, overlooking imperfections in reaction equations in popular datasets, such as missing reactants, products, other physical and practical constraints such as temperature and cost, primarily driven by a focus on the target molecule. This limitation leads to an incomplete representation of viable synthetic routes, especially when multiple sets of reactants can yield a given desired product. In response to these shortcomings, this study examines the prevailing evaluation methods and introduces comprehensive metrics designed to address imperfections in the dataset. Our novel metrics not only assess absolute accuracy by comparing predicted outputs with ground truth but also introduce a nuanced evaluation approach. We provide scores for partial correctness and compute adjusted accuracy through graph matching, acknowledging the inherent complexities of retrosynthetic pathways. Additionally, we explore the impact of small molecular augmentations while preserving chemical properties and employ similarity matching to enhance the assessment of prediction quality. We introduce SynFormer, a sequence-to-sequence model tailored for SMILES representation. It incorporates architectural enhancements to the original transformer, effectively tackling the challenges of chemical reaction prediction. SynFormer achieves a top-1 accuracy of 53.2% on the USPTO-50k dataset, demonstrating an improvement over previous state-of-the-art language models while being more efficient and eliminating the need for pre-training.

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

Retrosynthesis prediction using an end-to-end graph generative architecture for molecular graph editing

Chemical-Reaction-Aware Molecule Representation Learning

Self-Supervised Contrastive Molecular Representation Learning with a Chemical Synthesis Knowledge Graph

Chemistry-informed molecular graph as reaction descriptor for machine-learned retrosynthesis planning

Data-Efficient Graph Grammar Learning for Molecular Generation

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

Chemical reaction enhanced graph learning for molecule representation

Bridging the Gap between Chemical Reaction Pretraining and Conditional Molecule Generation with a Unified Model

MolGrapher: Graph-based Visual Recognition of Chemical Structures

Chemical Reaction Enhanced Graph Learning for Molecule Representation.

Learning Graph Models for Retrosynthesis Prediction

Learning Chemical Reaction Representation with Reactant-Product Alignment

Dissecting Errors in Machine Learning for Retrosynthesis: A Granular Metric Framework and Transformer-Based Model for More Informative Predictions

Molecule-Edit Templates for Efficient and Accurate Retrosynthesis Prediction

Learning to Plan Chemical Syntheses

Retrosynthetic reaction prediction using neural sequence-to-sequence models

Chimera: Accurate retrosynthesis prediction by ensembling models with diverse inductive biases

MechRetro is a chemical-mechanism-driven graph learning framework for interpretable retrosynthesis prediction and pathway planning

Molecular Graph Enhanced Transformer for Retrosynthesis Prediction

A Graph to Graphs Framework for Retrosynthesis Prediction