Transfer learning across different photocatalytic organic reactions

Abstract While seasoned organic chemists can often predict suitable catalysts for new reactions based on their past experiences in other catalytic reactions, developing this ability is costly, laborious and time-consuming. Therefore, replicating this remarkable expertize of human researchers through...

Full description

Saved in:
Bibliographic Details
Main Authors: Naoki Noto, Ryuga Kunisada, Tabea Rohlfs, Manami Hayashi, Ryosuke Kojima, Olga García Mancheño, Takeshi Yanai, Susumu Saito
Format: Article
Language:English
Published: Nature Portfolio 2025-04-01
Series:Nature Communications
Online Access:https://doi.org/10.1038/s41467-025-58687-5
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1850201068646957056
author Naoki Noto
Ryuga Kunisada
Tabea Rohlfs
Manami Hayashi
Ryosuke Kojima
Olga García Mancheño
Takeshi Yanai
Susumu Saito
author_facet Naoki Noto
Ryuga Kunisada
Tabea Rohlfs
Manami Hayashi
Ryosuke Kojima
Olga García Mancheño
Takeshi Yanai
Susumu Saito
author_sort Naoki Noto
collection DOAJ
description Abstract While seasoned organic chemists can often predict suitable catalysts for new reactions based on their past experiences in other catalytic reactions, developing this ability is costly, laborious and time-consuming. Therefore, replicating this remarkable expertize of human researchers through machine learning (ML) is compelling, albeit that it remains highly challenging. Herein, we apply a domain-adaptation-based transfer-learning (TL) approach to photocatalysis. Despite being different reaction types, the knowledge of the catalytic behavior of organic photosensitizers (OPSs) from photocatalytic cross-coupling reactions is successfully transferred to ML for a [2+2] cycloaddition reaction, improving the prediction of the photocatalytic activity compared with conventional ML approaches. Furthermore, a satisfactory predictive performance is achieved by using only ten training data points. This experimentally readily accessible small dataset can also be used to identify effective OPSs for alkene photoisomerization, thereby showcasing the potential benefits of TL in catalyst exploration.
format Article
id doaj-art-e3f45b796d3e498fa0e0033c317663c8
institution OA Journals
issn 2041-1723
language English
publishDate 2025-04-01
publisher Nature Portfolio
record_format Article
series Nature Communications
spelling doaj-art-e3f45b796d3e498fa0e0033c317663c82025-08-20T02:12:07ZengNature PortfolioNature Communications2041-17232025-04-0116111110.1038/s41467-025-58687-5Transfer learning across different photocatalytic organic reactionsNaoki Noto0Ryuga Kunisada1Tabea Rohlfs2Manami Hayashi3Ryosuke Kojima4Olga García Mancheño5Takeshi Yanai6Susumu Saito7Integrated Research Consortium on Chemical Sciences (IRCCS), Nagoya UniversityGraduate School of Science, Nagoya UniversityOrganic Chemistry Institute, University of MünsterGraduate School of Science, Nagoya UniversityDepartment of Biomedical Data Intelligence, Graduate School of Medicine, Kyoto UniversityOrganic Chemistry Institute, University of MünsterGraduate School of Science, Nagoya UniversityIntegrated Research Consortium on Chemical Sciences (IRCCS), Nagoya UniversityAbstract While seasoned organic chemists can often predict suitable catalysts for new reactions based on their past experiences in other catalytic reactions, developing this ability is costly, laborious and time-consuming. Therefore, replicating this remarkable expertize of human researchers through machine learning (ML) is compelling, albeit that it remains highly challenging. Herein, we apply a domain-adaptation-based transfer-learning (TL) approach to photocatalysis. Despite being different reaction types, the knowledge of the catalytic behavior of organic photosensitizers (OPSs) from photocatalytic cross-coupling reactions is successfully transferred to ML for a [2+2] cycloaddition reaction, improving the prediction of the photocatalytic activity compared with conventional ML approaches. Furthermore, a satisfactory predictive performance is achieved by using only ten training data points. This experimentally readily accessible small dataset can also be used to identify effective OPSs for alkene photoisomerization, thereby showcasing the potential benefits of TL in catalyst exploration.https://doi.org/10.1038/s41467-025-58687-5
spellingShingle Naoki Noto
Ryuga Kunisada
Tabea Rohlfs
Manami Hayashi
Ryosuke Kojima
Olga García Mancheño
Takeshi Yanai
Susumu Saito
Transfer learning across different photocatalytic organic reactions
Nature Communications
title Transfer learning across different photocatalytic organic reactions
title_full Transfer learning across different photocatalytic organic reactions
title_fullStr Transfer learning across different photocatalytic organic reactions
title_full_unstemmed Transfer learning across different photocatalytic organic reactions
title_short Transfer learning across different photocatalytic organic reactions
title_sort transfer learning across different photocatalytic organic reactions
url https://doi.org/10.1038/s41467-025-58687-5
work_keys_str_mv AT naokinoto transferlearningacrossdifferentphotocatalyticorganicreactions
AT ryugakunisada transferlearningacrossdifferentphotocatalyticorganicreactions
AT tabearohlfs transferlearningacrossdifferentphotocatalyticorganicreactions
AT manamihayashi transferlearningacrossdifferentphotocatalyticorganicreactions
AT ryosukekojima transferlearningacrossdifferentphotocatalyticorganicreactions
AT olgagarciamancheno transferlearningacrossdifferentphotocatalyticorganicreactions
AT takeshiyanai transferlearningacrossdifferentphotocatalyticorganicreactions
AT susumusaito transferlearningacrossdifferentphotocatalyticorganicreactions