AiZynthTrain: robust, reproducible, and extensible pipelines for training synthesis prediction models



We introduce the AiZynthTrain Python package for training synthesis models in a robust, reproducible, and extensible way. It contains two pipelines that create a template-based one-step retrosynthesis model and a RingBreaker model that can be straightforwardly integrated in retrosynthesis software. We train such models on the publicly available reaction dataset from the US Patent and Trademark Office (USPTO), and these are the first retrosynthesis models created in a completely reproducible end-to-end fashion, starting with the original reaction data source and ending with trained machine-learning models. In particular, we show that employing the pipeline greatly improves the ability of the RingBreaker model for disconnecting ring systems. Furthermore, we demonstrate the robustness of the pipeline by training on a more diverse but proprietary dataset. We envisage that this framework will be extended with other synthesis models in the future.


Supplementary material

Pipeline reports for USPTO
These are 4 reports generated by the pipelines for the USPTO-based models

Supplementary weblinks

Code repository
The GitHub code repository
Models and datasets
Publicly released models and datasets created in this paper