{"about":{"site":"https://codewithpapers.app","non_affiliation":"Code with Papers and Syntology are not affiliated with, endorsed by, or sponsored by Papers with Code, Meta, or the pwc-archive mirror.","licence":"CC BY-SA 4.0","licence_url":"https://creativecommons.org/licenses/by-sa/4.0/legalcode","attribution":"https://codewithpapers.app/attribution","modified":"archive material modified by Syntology; see the attribution page"},"url":"/paper/retrosynthetic-reaction-prediction-using","title":"Retrosynthetic reaction prediction using neural sequence-to-sequence models","arxiv_id":"1706.01643","date":"2017-06-06","proceeding":null,"authors":["Bowen Liu","Bharath Ramsundar","Prasad Kawthekar","Jade Shi","Joseph Gomes","Quang Luu Nguyen","Stephen Ho","Jack Sloane","Paul Wender","Vijay Pande"],"abstract":"We describe a fully data driven model that learns to perform a retrosynthetic\nreaction prediction task, which is treated as a sequence-to-sequence mapping\nproblem. The end-to-end trained model has an encoder-decoder architecture that\nconsists of two recurrent neural networks, which has previously shown great\nsuccess in solving other sequence-to-sequence prediction tasks such as machine\ntranslation. The model is trained on 50,000 experimental reaction examples from\nthe United States patent literature, which span 10 broad reaction types that\nare commonly used by medicinal chemists. We find that our model performs\ncomparably with a rule-based expert system baseline model, and also overcomes\ncertain limitations associated with rule-based expert systems and with any\nmachine learning approach that contains a rule-based expert system component.\nOur model provides an important first step towards solving the challenging\nproblem of computational retrosynthetic analysis.","url_abs":"http://arxiv.org/abs/1706.01643v1","url_pdf":"http://arxiv.org/pdf/1706.01643v1.pdf","source":{"archive":"pwc-archive (Hugging Face), CC BY-SA 4.0","snapshot":"2025-07-28","licence_url":"https://creativecommons.org/licenses/by-sa/4.0/legalcode","row_kind":"abstracts"},"code_links":[{"paper_slug":"retrosynthetic-reaction-prediction-using","repo_url":"https://github.com/pandegroup/reaction_prediction_seq2seq","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":0,"framework":"tf","reach":null}],"tasks":[{"task_slug":"decoder","task_name":"Decoder"},{"task_slug":"machine-translation","task_name":"Machine Translation"},{"task_slug":"prediction","task_name":"Prediction"},{"task_slug":"single-step-retrosynthesis","task_name":"Single-step retrosynthesis"},{"task_slug":"translation","task_name":"Translation"}],"methods":[],"datasets_introduced":[],"methods_introduced":[],"results":[{"leaderboard":"/sota/single-step-retrosynthesis-on-uspto-50k","task":"Single-step retrosynthesis","dataset":"USPTO-50k","model":"Expert System (reaction class as prior)","rank_in_archive_order":35,"of":35,"metrics":{"Top-1 accuracy":"35.2","Top-10 accuracy":"65.1","Top-3 accuracy":"52.3","Top-5 accuracy":"59.1"},"uses_additional_data":false}],"syntology":{"atlas_url":"https://app.syntology.ai/?focus=1706.01643","mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}