RetroMPA: A Molecular Property-Aware Auxiliary Framework for Enhancing Retrosynthesis Prediction
2026-08-17 • Machine Learning
Machine LearningArtificial Intelligence
AI summaryⓘ
The authors developed RetroMPA, a tool that improves drug-making recipe predictions by adding chemical knowledge to existing AI models without changing their design or needing huge retraining efforts. Instead of creating new suggestions from scratch, RetroMPA acts like a smart filter that makes current predictions better. They tested it on popular datasets and found it consistently improved prediction accuracy. They also did lab experiments confirming some new, workable chemical reactions found using the tool, showing it can help beyond just studying data.
RetrosynthesisDeep learningMolecular propertySMILESUSPTO datasetTemplate-based methodsTemplate-free methodsSuzuki-Miyaura couplingBucherer reactionFriedel-Crafts acylation
Authors
Mianzhi Liu, Fan Xiao, Zhiliang Yu, Huayang Huang, Yuke Li, Yi Yang, Wenbo Liu, Yu Wu
Abstract
Retrosynthesis is a cornerstone of drug discovery and organic synthesis. While data-driven deep learning models have shown remarkable progress, they autonomously learn reaction patterns from extensive datasets with limited integration of established chemical knowledge as priors. To address this limitation, we introduce RetroMPA, a molecular property-aware, post-hoc enhancement module that injects chemical knowledge into the retrosynthesis pipeline. Rather than functioning as an independent SMILES sequence generator, RetroMPA is a broadly applicable, model-agnostic chemical filter designed to recalibrate and optimize the predictive pathways of existing algorithms. This plug-and-play framework integrates seamlessly with a range of data-driven retrosynthesis methods, enhancing outputs without modifying model architecture or requiring resource-intensive retraining. By leveraging a property-aware latent embedding space, RetroMPA consistently improves top-1 accuracy across eight representative retrosynthesis models by an average of 5.50% on USPTO-50K. Furthermore, we validate its scalability on the large-scale USPTO-Full dataset, achieving an average improvement of about 2.03% across both template-based and template-free architectures. Wet-lab experiments provide preliminary support for the practical utility of the framework. These syntheses confirmed viable, previously unreported substrate combinations for classic reaction paradigms---specifically, Suzuki-Miyaura coupling, Bucherer reaction, and Friedel-Crafts acylation---suggesting that RetroMPA can operate beyond mere data fitting. The code is open-sourced at https://github.com/MengzhouLu/RetroMPA.