$$\texttt {DiffER}$$ DiffER : categorical diffusion ensembles for single-step chemical retrosynthesis
Abstract Methods for automatic chemical retrosynthesis have found recent success through the application of models traditionally built for natural language processing, primarily through transformer neural networks. These models have demonstrated significant ability to translate between the SMILES en...
Saved in:
| Main Authors: | , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
BMC
2025-07-01
|
| Series: | Journal of Cheminformatics |
| Online Access: | https://doi.org/10.1186/s13321-025-01056-7 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1849761348054941696 |
|---|---|
| author | Sean Current Ziqi Chen Daniel Adu-Ampratwum Xia Ning Srinivasan Parthasarathy |
| author_facet | Sean Current Ziqi Chen Daniel Adu-Ampratwum Xia Ning Srinivasan Parthasarathy |
| author_sort | Sean Current |
| collection | DOAJ |
| description | Abstract Methods for automatic chemical retrosynthesis have found recent success through the application of models traditionally built for natural language processing, primarily through transformer neural networks. These models have demonstrated significant ability to translate between the SMILES encodings of chemical products and reactants, but are constrained as a result of their autoregressive nature. We propose $$\texttt {DiffER}$$ DiffER , an alternative template-free method for single-step retrosynthesis prediction in the form of categorical diffusion, which allows the entire output SMILES sequence to be predicted in unison. We construct an ensemble of diffusion models which achieves state-of-the-art performance for top-1 accuracy and competitive performance for top-3, top-5, and top-10 accuracy among template-free methods. We prove that $$\texttt {DiffER}$$ DiffER is a strong baseline for a new class of template-free model and is capable of learning a variety of synthetic techniques used in laboratory settings. |
| format | Article |
| id | doaj-art-bafa542ff7ac450aa66597e4fdf10ba0 |
| institution | DOAJ |
| issn | 1758-2946 |
| language | English |
| publishDate | 2025-07-01 |
| publisher | BMC |
| record_format | Article |
| series | Journal of Cheminformatics |
| spelling | doaj-art-bafa542ff7ac450aa66597e4fdf10ba02025-08-20T03:06:04ZengBMCJournal of Cheminformatics1758-29462025-07-0117111610.1186/s13321-025-01056-7$$\texttt {DiffER}$$ DiffER : categorical diffusion ensembles for single-step chemical retrosynthesisSean Current0Ziqi Chen1Daniel Adu-Ampratwum2Xia Ning3Srinivasan Parthasarathy4Computer Science and Engineering, The Ohio State UniversityComputer Science and Engineering, The Ohio State UniversityCollege of Pharmacy, The Ohio State UniversityComputer Science and Engineering, The Ohio State UniversityComputer Science and Engineering, The Ohio State UniversityAbstract Methods for automatic chemical retrosynthesis have found recent success through the application of models traditionally built for natural language processing, primarily through transformer neural networks. These models have demonstrated significant ability to translate between the SMILES encodings of chemical products and reactants, but are constrained as a result of their autoregressive nature. We propose $$\texttt {DiffER}$$ DiffER , an alternative template-free method for single-step retrosynthesis prediction in the form of categorical diffusion, which allows the entire output SMILES sequence to be predicted in unison. We construct an ensemble of diffusion models which achieves state-of-the-art performance for top-1 accuracy and competitive performance for top-3, top-5, and top-10 accuracy among template-free methods. We prove that $$\texttt {DiffER}$$ DiffER is a strong baseline for a new class of template-free model and is capable of learning a variety of synthetic techniques used in laboratory settings.https://doi.org/10.1186/s13321-025-01056-7 |
| spellingShingle | Sean Current Ziqi Chen Daniel Adu-Ampratwum Xia Ning Srinivasan Parthasarathy $$\texttt {DiffER}$$ DiffER : categorical diffusion ensembles for single-step chemical retrosynthesis Journal of Cheminformatics |
| title | $$\texttt {DiffER}$$ DiffER : categorical diffusion ensembles for single-step chemical retrosynthesis |
| title_full | $$\texttt {DiffER}$$ DiffER : categorical diffusion ensembles for single-step chemical retrosynthesis |
| title_fullStr | $$\texttt {DiffER}$$ DiffER : categorical diffusion ensembles for single-step chemical retrosynthesis |
| title_full_unstemmed | $$\texttt {DiffER}$$ DiffER : categorical diffusion ensembles for single-step chemical retrosynthesis |
| title_short | $$\texttt {DiffER}$$ DiffER : categorical diffusion ensembles for single-step chemical retrosynthesis |
| title_sort | texttt differ differ categorical diffusion ensembles for single step chemical retrosynthesis |
| url | https://doi.org/10.1186/s13321-025-01056-7 |
| work_keys_str_mv | AT seancurrent textttdifferdiffercategoricaldiffusionensemblesforsinglestepchemicalretrosynthesis AT ziqichen textttdifferdiffercategoricaldiffusionensemblesforsinglestepchemicalretrosynthesis AT danieladuampratwum textttdifferdiffercategoricaldiffusionensemblesforsinglestepchemicalretrosynthesis AT xianing textttdifferdiffercategoricaldiffusionensemblesforsinglestepchemicalretrosynthesis AT srinivasanparthasarathy textttdifferdiffercategoricaldiffusionensemblesforsinglestepchemicalretrosynthesis |