Saved in:
| Main Authors: | , , , , , , , , , |
|---|---|
| Format: | Preprint |
| Published: |
2025
|
| Subjects: | |
| Online Access: | https://arxiv.org/abs/2510.10645 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866908697983188992 |
|---|---|
| author | Sadowski, Michal Radusinović, Tadija Wyrzykowska, Maria Sztukiewicz, Lukasz Rzymkowski, Jan Włodarczyk-Pruszyński, Paweł Sacha, Mikołaj Kozakowski, Piotr van Workum, Ruard Jastrzebski, Stanislaw Kamil |
| author_facet | Sadowski, Michal Radusinović, Tadija Wyrzykowska, Maria Sztukiewicz, Lukasz Rzymkowski, Jan Włodarczyk-Pruszyński, Paweł Sacha, Mikołaj Kozakowski, Piotr van Workum, Ruard Jastrzebski, Stanislaw Kamil |
| contents | Retrosynthesis is one of the domains transformed by the rise of generative models, and it is one where the problem of nonsensical or erroneous outputs (hallucinations) is particularly insidious: reliable assessment of synthetic plans is time-consuming, with automatic methods lacking. In this work, we present RetroTrim, a retrosynthesis system that successfully avoids nonsensical plans on a set of challenging drug-like targets. Compared to common baselines in the field, our system is not only the sole method that succeeds in filtering out hallucinated reactions, but it also results in the highest number of high-quality paths overall. The key insight behind RetroTrim is the combination of diverse reaction scoring strategies, based on machine learning models and existing chemical databases. We show that our scoring strategies capture different classes of hallucinations by analyzing them on a dataset of labeled retrosynthetic intermediates. This approach formed the basis of our winning solution to the Standard Industries \$1 million Retrosynthesis Challenge. To measure the performance of retrosynthesis systems, we propose a novel evaluation protocol for reactions and synthetic paths based on a structured review by expert chemists. Using this protocol, we compare systems on a set of 32 novel targets, curated to reflect recent trends in drug structures. While the insights behind our methodology are broadly applicable to retrosynthesis, our focus is on targets in the drug-like domain. By releasing our benchmark targets and the details of our evaluation protocol, we hope to inspire further research into reliable retrosynthesis. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2510_10645 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | Trustworthy Retrosynthesis: Eliminating Hallucinations with a Diverse Ensemble of Reaction Scorers Sadowski, Michal Radusinović, Tadija Wyrzykowska, Maria Sztukiewicz, Lukasz Rzymkowski, Jan Włodarczyk-Pruszyński, Paweł Sacha, Mikołaj Kozakowski, Piotr van Workum, Ruard Jastrzebski, Stanislaw Kamil Machine Learning Artificial Intelligence Retrosynthesis is one of the domains transformed by the rise of generative models, and it is one where the problem of nonsensical or erroneous outputs (hallucinations) is particularly insidious: reliable assessment of synthetic plans is time-consuming, with automatic methods lacking. In this work, we present RetroTrim, a retrosynthesis system that successfully avoids nonsensical plans on a set of challenging drug-like targets. Compared to common baselines in the field, our system is not only the sole method that succeeds in filtering out hallucinated reactions, but it also results in the highest number of high-quality paths overall. The key insight behind RetroTrim is the combination of diverse reaction scoring strategies, based on machine learning models and existing chemical databases. We show that our scoring strategies capture different classes of hallucinations by analyzing them on a dataset of labeled retrosynthetic intermediates. This approach formed the basis of our winning solution to the Standard Industries \$1 million Retrosynthesis Challenge. To measure the performance of retrosynthesis systems, we propose a novel evaluation protocol for reactions and synthetic paths based on a structured review by expert chemists. Using this protocol, we compare systems on a set of 32 novel targets, curated to reflect recent trends in drug structures. While the insights behind our methodology are broadly applicable to retrosynthesis, our focus is on targets in the drug-like domain. By releasing our benchmark targets and the details of our evaluation protocol, we hope to inspire further research into reliable retrosynthesis. |
| title | Trustworthy Retrosynthesis: Eliminating Hallucinations with a Diverse Ensemble of Reaction Scorers |
| topic | Machine Learning Artificial Intelligence |
| url | https://arxiv.org/abs/2510.10645 |