Retrosynthesis
42 papers with code • 1 benchmarks • 0 datasets
Retrosynthetic analysis is a pivotal synthetic methodology in organic chemistry that employs a reverse-engineering approach, initiating from the target compound and retroactively tracing potential synthesis routes and precursor molecules. This technique proves instrumental in sculpting efficient synthetic strategies for intricate molecules, thus catalyzing the evolution and progression of novel pharmaceuticals and materials.
Benchmarks
These leaderboards are used to track progress in Retrosynthesis
| Trend | Dataset | Best Model | Paper | Code | Compare |
|---|---|---|---|---|---|
|
Most implemented papers
Deep Retrosynthetic Reaction Prediction using Local Reactivity and Global Attention
kaist-amsg/LocalRetro
•
•
JACS Au 2021
Our model 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.
RetroPrime: A Diverse, plausible and Transformer-based method for Single-Step retrosynthesis predictions
wangxr0526/RetroPrime
•
•
Chemical Engineering Journal 2021
RetroPrime achieves the Top-1 accuracy of 64. 8% and 51. 4%, when the reaction type is known and unknown, respectively, in the USPTO-50 K dataset.
Motif-based Graph Self-Supervised Learning for Molecular Property Prediction
zaixizhang/MGSSL
•
•
NeurIPS 2021
To bridge this gap, we propose Motif-based Graph Self-supervised Learning (MGSSL) by introducing a novel self-supervised motif generation framework for GNNs.
Permutation invariant graph-to-sequence model for template-free retrosynthesis and reaction prediction
Synthesis planning and reaction outcome prediction are two fundamental problems in computer-aided organic chemistry for which a variety of data-driven approaches have emerged.
RetroComposer: Composing Templates for Template-Based Retrosynthesis Prediction
To overcome this limitation, we propose an innovative retrosynthesis prediction framework that can compose novel templates beyond training templates.
Improving Few- and Zero-Shot Reaction Template Prediction Using Modern Hopfield Networks
ml-jku/mhn-react
•
•
Journal of Chemical Information and Modeling 2022
Finding synthesis routes for molecules of interest is essential in the discovery of new drugs and materials.
Retroformer: Pushing the Limits of Interpretable End-to-end Retrosynthesis Transformer
In this paper, we propose Retroformer, a novel Transformer-based architecture for retrosynthesis prediction without relying on any cheminformatics tools for molecule editing.
Chemformer: a pre-trained transformer for computational chemistry
MolecularAI/Chemformer
•
•
Machine Learning: Science and Technology 2022
We also improve on existing approaches for a molecular optimisation task and show that Chemformer can optimise on multiple discriminative tasks simultaneously.
Root-aligned SMILES: A Tight Representation for Chemical Reaction Prediction
Chemical reaction prediction, involving forward synthesis and retrosynthesis prediction, is a fundamental problem in organic synthesis.
Leveraging Reaction-aware Substructures for Retrosynthesis Analysis
In this paper, we propose a substructure-level decoding model, where the substructures are reaction-aware and can be automatically extracted with a fully data-driven approach.
