Search papers, labs, and topics across Lattice.
This paper introduces RetroMPA, a molecular property-aware enhancement module that integrates chemical knowledge into existing retrosynthesis prediction models without altering their architecture. By utilizing a property-aware latent embedding space, RetroMPA improves the top-1 accuracy of eight retrosynthesis models by an average of 5.50% on the USPTO-50K dataset and shows a 2.03% improvement on the larger USPTO-Full dataset. Preliminary wet-lab experiments validate its practical utility by revealing viable substrate combinations for established reaction paradigms, indicating its potential to enhance synthetic pathways beyond data fitting.
RetroMPA boosts retrosynthesis prediction accuracy by injecting chemical knowledge into existing models, achieving significant improvements without the need for retraining.
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.
