Loading…
Identifying general reaction conditions by bandit optimization
Reaction conditions that are generally applicable to a wide variety of substrates are highly desired, especially in the pharmaceutical and chemical industries . Although many approaches are available to evaluate the general applicability of developed conditions, a universal approach to efficiently d...
Saved in:
| Published in: | Nature (London) 2024-02, Vol.626 (8001), p.1025-1033 |
|---|---|
| Main Authors: | , , , , , , , , , , , , , , , , , |
| Format: | Article |
| Language: | English |
| Subjects: | |
| Citations: | Items that this one cites |
| Online Access: | Get full text |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| cited_by | |
|---|---|
| cites | cdi_FETCH-LOGICAL-c282t-ea04ad522e85a51fb08f51135756ab76cbfda1214df04350cc6006cc6fa8d3ff3 |
| container_end_page | 1033 |
| container_issue | 8001 |
| container_start_page | 1025 |
| container_title | Nature (London) |
| container_volume | 626 |
| creator | Wang, Jason Y Stevens, Jason M Kariofillis, Stavros K Tom, Mai-Jan Golden, Dung L Li, Jun Tabora, Jose E Parasram, Marvin Shields, Benjamin J Primer, David N Hao, Bo Del Valle, David DiSomma, Stacey Furman, Ariel Zipp, G Greg Melnikov, Sergey Paulson, James Doyle, Abigail G |
| description | Reaction conditions that are generally applicable to a wide variety of substrates are highly desired, especially in the pharmaceutical and chemical industries
. Although many approaches are available to evaluate the general applicability of developed conditions, a universal approach to efficiently discover these conditions during optimizations is rare. Here we report the design, implementation and application of reinforcement learning bandit optimization models
to identify generally applicable conditions by efficient condition sampling and evaluation of experimental feedback. Performance benchmarking on existing datasets statistically showed high accuracies for identifying general conditions, with up to 31% improvement over baselines that mimic state-of-the-art optimization approaches. A palladium-catalysed imidazole C-H arylation reaction, an aniline amide coupling reaction and a phenol alkylation reaction were investigated experimentally to evaluate use cases and functionalities of the bandit optimization model in practice. In all three cases, the reaction conditions that were most generally applicable yet not well studied for the respective reaction were identified after surveying less than 15% of the expert-designed reaction space. |
| doi_str_mv | 10.1038/s41586-024-07021-y |
| format | article |
| fullrecord | <record><control><sourceid>proquest_cross</sourceid><recordid>TN_cdi_proquest_miscellaneous_2933462629</recordid><sourceformat>XML</sourceformat><sourcesystem>PC</sourcesystem><sourcerecordid>2937395795</sourcerecordid><originalsourceid>FETCH-LOGICAL-c282t-ea04ad522e85a51fb08f51135756ab76cbfda1214df04350cc6006cc6fa8d3ff3</originalsourceid><addsrcrecordid>eNpdkE1LAzEQhoMoWqt_wIMsePESnXxnL4IUPwoFL3oO2d2kpOxHTbaH9de7a6sHL_PODO-8DA9CVwTuCDB9nzgRWmKgHIMCSvBwhGaEK4m51OoYzQCoxqCZPEPnKW0AQBDFT9EZ05zonNAZelhWru2DH0K7ztauddHWWXS27EPXZmXXVmHqUlYMWWGnKeu2fWjCl532F-jE2zq5y4PO0cfz0_viFa_eXpaLxxUuqaY9dha4rQSlTgsriC9Ae0EIE0pIWyhZFr6yhBJeeeBMQFlKADlWb3XFvGdzdLvP3cbuc-dSb5qQSlfXtnXdLhmaM8YllaPO0c0_66bbxXb8bnIplguVi9FF964ydilF5802hsbGwRAwE12zp2tGuuaHrhnGo-tD9K5oXPV38ouTfQMrqnV1</addsrcrecordid><sourcetype>Aggregation Database</sourcetype><iscdi>true</iscdi><recordtype>article</recordtype><pqid>2937395795</pqid></control><display><type>article</type><title>Identifying general reaction conditions by bandit optimization</title><source>Springer Nature - Connect here FIRST to enable access</source><creator>Wang, Jason Y ; Stevens, Jason M ; Kariofillis, Stavros K ; Tom, Mai-Jan ; Golden, Dung L ; Li, Jun ; Tabora, Jose E ; Parasram, Marvin ; Shields, Benjamin J ; Primer, David N ; Hao, Bo ; Del Valle, David ; DiSomma, Stacey ; Furman, Ariel ; Zipp, G Greg ; Melnikov, Sergey ; Paulson, James ; Doyle, Abigail G</creator><creatorcontrib>Wang, Jason Y ; Stevens, Jason M ; Kariofillis, Stavros K ; Tom, Mai-Jan ; Golden, Dung L ; Li, Jun ; Tabora, Jose E ; Parasram, Marvin ; Shields, Benjamin J ; Primer, David N ; Hao, Bo ; Del Valle, David ; DiSomma, Stacey ; Furman, Ariel ; Zipp, G Greg ; Melnikov, Sergey ; Paulson, James ; Doyle, Abigail G</creatorcontrib><description>Reaction conditions that are generally applicable to a wide variety of substrates are highly desired, especially in the pharmaceutical and chemical industries
. Although many approaches are available to evaluate the general applicability of developed conditions, a universal approach to efficiently discover these conditions during optimizations is rare. Here we report the design, implementation and application of reinforcement learning bandit optimization models
to identify generally applicable conditions by efficient condition sampling and evaluation of experimental feedback. Performance benchmarking on existing datasets statistically showed high accuracies for identifying general conditions, with up to 31% improvement over baselines that mimic state-of-the-art optimization approaches. A palladium-catalysed imidazole C-H arylation reaction, an aniline amide coupling reaction and a phenol alkylation reaction were investigated experimentally to evaluate use cases and functionalities of the bandit optimization model in practice. In all three cases, the reaction conditions that were most generally applicable yet not well studied for the respective reaction were identified after surveying less than 15% of the expert-designed reaction space.</description><identifier>ISSN: 0028-0836</identifier><identifier>EISSN: 1476-4687</identifier><identifier>DOI: 10.1038/s41586-024-07021-y</identifier><identifier>PMID: 38418912</identifier><language>eng</language><publisher>England: Nature Publishing Group</publisher><subject>Algorithms ; Alkylation ; Aniline ; Chemical reactions ; Chemistry ; Chemists ; Datasets ; Design ; Gaming machines ; Imidazole ; Optimization ; Optimization models ; Palladium ; Phenols ; Software ; Substrates</subject><ispartof>Nature (London), 2024-02, Vol.626 (8001), p.1025-1033</ispartof><rights>2024. The Author(s), under exclusive licence to Springer Nature Limited.</rights><rights>Copyright Nature Publishing Group Feb 29, 2024</rights><lds50>peer_reviewed</lds50><woscitedreferencessubscribed>false</woscitedreferencessubscribed><cites>FETCH-LOGICAL-c282t-ea04ad522e85a51fb08f51135756ab76cbfda1214df04350cc6006cc6fa8d3ff3</cites><orcidid>0000-0003-1764-4497 ; 0000-0002-5461-3190 ; 0000-0002-6641-0833 ; 0000-0003-1671-1539 ; 0000-0002-9044-4873 ; 0000-0002-0594-7143 ; 0000-0001-5826-2554 ; 0009-0000-3043-9637</orcidid></display><links><openurl>$$Topenurl_article</openurl><openurlfulltext>$$Topenurlfull_article</openurlfulltext><thumbnail>$$Tsyndetics_thumb_exl</thumbnail><link.rule.ids>315,786,790,27957,27958</link.rule.ids><backlink>$$Uhttps://www.ncbi.nlm.nih.gov/pubmed/38418912$$D View this record in MEDLINE/PubMed$$Hfree_for_read</backlink></links><search><creatorcontrib>Wang, Jason Y</creatorcontrib><creatorcontrib>Stevens, Jason M</creatorcontrib><creatorcontrib>Kariofillis, Stavros K</creatorcontrib><creatorcontrib>Tom, Mai-Jan</creatorcontrib><creatorcontrib>Golden, Dung L</creatorcontrib><creatorcontrib>Li, Jun</creatorcontrib><creatorcontrib>Tabora, Jose E</creatorcontrib><creatorcontrib>Parasram, Marvin</creatorcontrib><creatorcontrib>Shields, Benjamin J</creatorcontrib><creatorcontrib>Primer, David N</creatorcontrib><creatorcontrib>Hao, Bo</creatorcontrib><creatorcontrib>Del Valle, David</creatorcontrib><creatorcontrib>DiSomma, Stacey</creatorcontrib><creatorcontrib>Furman, Ariel</creatorcontrib><creatorcontrib>Zipp, G Greg</creatorcontrib><creatorcontrib>Melnikov, Sergey</creatorcontrib><creatorcontrib>Paulson, James</creatorcontrib><creatorcontrib>Doyle, Abigail G</creatorcontrib><title>Identifying general reaction conditions by bandit optimization</title><title>Nature (London)</title><addtitle>Nature</addtitle><description>Reaction conditions that are generally applicable to a wide variety of substrates are highly desired, especially in the pharmaceutical and chemical industries
. Although many approaches are available to evaluate the general applicability of developed conditions, a universal approach to efficiently discover these conditions during optimizations is rare. Here we report the design, implementation and application of reinforcement learning bandit optimization models
to identify generally applicable conditions by efficient condition sampling and evaluation of experimental feedback. Performance benchmarking on existing datasets statistically showed high accuracies for identifying general conditions, with up to 31% improvement over baselines that mimic state-of-the-art optimization approaches. A palladium-catalysed imidazole C-H arylation reaction, an aniline amide coupling reaction and a phenol alkylation reaction were investigated experimentally to evaluate use cases and functionalities of the bandit optimization model in practice. In all three cases, the reaction conditions that were most generally applicable yet not well studied for the respective reaction were identified after surveying less than 15% of the expert-designed reaction space.</description><subject>Algorithms</subject><subject>Alkylation</subject><subject>Aniline</subject><subject>Chemical reactions</subject><subject>Chemistry</subject><subject>Chemists</subject><subject>Datasets</subject><subject>Design</subject><subject>Gaming machines</subject><subject>Imidazole</subject><subject>Optimization</subject><subject>Optimization models</subject><subject>Palladium</subject><subject>Phenols</subject><subject>Software</subject><subject>Substrates</subject><issn>0028-0836</issn><issn>1476-4687</issn><fulltext>true</fulltext><rsrctype>article</rsrctype><creationdate>2024</creationdate><recordtype>article</recordtype><recordid>eNpdkE1LAzEQhoMoWqt_wIMsePESnXxnL4IUPwoFL3oO2d2kpOxHTbaH9de7a6sHL_PODO-8DA9CVwTuCDB9nzgRWmKgHIMCSvBwhGaEK4m51OoYzQCoxqCZPEPnKW0AQBDFT9EZ05zonNAZelhWru2DH0K7ztauddHWWXS27EPXZmXXVmHqUlYMWWGnKeu2fWjCl532F-jE2zq5y4PO0cfz0_viFa_eXpaLxxUuqaY9dha4rQSlTgsriC9Ae0EIE0pIWyhZFr6yhBJeeeBMQFlKADlWb3XFvGdzdLvP3cbuc-dSb5qQSlfXtnXdLhmaM8YllaPO0c0_66bbxXb8bnIplguVi9FF964ydilF5802hsbGwRAwE12zp2tGuuaHrhnGo-tD9K5oXPV38ouTfQMrqnV1</recordid><startdate>20240229</startdate><enddate>20240229</enddate><creator>Wang, Jason Y</creator><creator>Stevens, Jason M</creator><creator>Kariofillis, Stavros K</creator><creator>Tom, Mai-Jan</creator><creator>Golden, Dung L</creator><creator>Li, Jun</creator><creator>Tabora, Jose E</creator><creator>Parasram, Marvin</creator><creator>Shields, Benjamin J</creator><creator>Primer, David N</creator><creator>Hao, Bo</creator><creator>Del Valle, David</creator><creator>DiSomma, Stacey</creator><creator>Furman, Ariel</creator><creator>Zipp, G Greg</creator><creator>Melnikov, Sergey</creator><creator>Paulson, James</creator><creator>Doyle, Abigail G</creator><general>Nature Publishing Group</general><scope>NPM</scope><scope>AAYXX</scope><scope>CITATION</scope><scope>7QG</scope><scope>7QL</scope><scope>7QP</scope><scope>7QR</scope><scope>7SN</scope><scope>7SS</scope><scope>7ST</scope><scope>7T5</scope><scope>7TG</scope><scope>7TK</scope><scope>7TM</scope><scope>7TO</scope><scope>7U9</scope><scope>8FD</scope><scope>C1K</scope><scope>FR3</scope><scope>H94</scope><scope>K9.</scope><scope>KL.</scope><scope>M7N</scope><scope>NAPCQ</scope><scope>P64</scope><scope>RC3</scope><scope>SOI</scope><scope>7X8</scope><orcidid>https://orcid.org/0000-0003-1764-4497</orcidid><orcidid>https://orcid.org/0000-0002-5461-3190</orcidid><orcidid>https://orcid.org/0000-0002-6641-0833</orcidid><orcidid>https://orcid.org/0000-0003-1671-1539</orcidid><orcidid>https://orcid.org/0000-0002-9044-4873</orcidid><orcidid>https://orcid.org/0000-0002-0594-7143</orcidid><orcidid>https://orcid.org/0000-0001-5826-2554</orcidid><orcidid>https://orcid.org/0009-0000-3043-9637</orcidid></search><sort><creationdate>20240229</creationdate><title>Identifying general reaction conditions by bandit optimization</title><author>Wang, Jason Y ; Stevens, Jason M ; Kariofillis, Stavros K ; Tom, Mai-Jan ; Golden, Dung L ; Li, Jun ; Tabora, Jose E ; Parasram, Marvin ; Shields, Benjamin J ; Primer, David N ; Hao, Bo ; Del Valle, David ; DiSomma, Stacey ; Furman, Ariel ; Zipp, G Greg ; Melnikov, Sergey ; Paulson, James ; Doyle, Abigail G</author></sort><facets><frbrtype>5</frbrtype><frbrgroupid>cdi_FETCH-LOGICAL-c282t-ea04ad522e85a51fb08f51135756ab76cbfda1214df04350cc6006cc6fa8d3ff3</frbrgroupid><rsrctype>articles</rsrctype><prefilter>articles</prefilter><language>eng</language><creationdate>2024</creationdate><topic>Algorithms</topic><topic>Alkylation</topic><topic>Aniline</topic><topic>Chemical reactions</topic><topic>Chemistry</topic><topic>Chemists</topic><topic>Datasets</topic><topic>Design</topic><topic>Gaming machines</topic><topic>Imidazole</topic><topic>Optimization</topic><topic>Optimization models</topic><topic>Palladium</topic><topic>Phenols</topic><topic>Software</topic><topic>Substrates</topic><toplevel>peer_reviewed</toplevel><toplevel>online_resources</toplevel><creatorcontrib>Wang, Jason Y</creatorcontrib><creatorcontrib>Stevens, Jason M</creatorcontrib><creatorcontrib>Kariofillis, Stavros K</creatorcontrib><creatorcontrib>Tom, Mai-Jan</creatorcontrib><creatorcontrib>Golden, Dung L</creatorcontrib><creatorcontrib>Li, Jun</creatorcontrib><creatorcontrib>Tabora, Jose E</creatorcontrib><creatorcontrib>Parasram, Marvin</creatorcontrib><creatorcontrib>Shields, Benjamin J</creatorcontrib><creatorcontrib>Primer, David N</creatorcontrib><creatorcontrib>Hao, Bo</creatorcontrib><creatorcontrib>Del Valle, David</creatorcontrib><creatorcontrib>DiSomma, Stacey</creatorcontrib><creatorcontrib>Furman, Ariel</creatorcontrib><creatorcontrib>Zipp, G Greg</creatorcontrib><creatorcontrib>Melnikov, Sergey</creatorcontrib><creatorcontrib>Paulson, James</creatorcontrib><creatorcontrib>Doyle, Abigail G</creatorcontrib><collection>PubMed</collection><collection>CrossRef</collection><collection>Animal Behavior Abstracts</collection><collection>Bacteriology Abstracts (Microbiology B)</collection><collection>Calcium & Calcified Tissue Abstracts</collection><collection>Chemoreception Abstracts</collection><collection>Ecology Abstracts</collection><collection>Entomology Abstracts (Full archive)</collection><collection>Environment Abstracts</collection><collection>Immunology Abstracts</collection><collection>Meteorological & Geoastrophysical Abstracts</collection><collection>Neurosciences Abstracts</collection><collection>Nucleic Acids Abstracts</collection><collection>Oncogenes and Growth Factors Abstracts</collection><collection>Virology and AIDS Abstracts</collection><collection>Technology Research Database</collection><collection>Environmental Sciences and Pollution Management</collection><collection>Engineering Research Database</collection><collection>AIDS and Cancer Research Abstracts</collection><collection>ProQuest Health & Medical Complete (Alumni)</collection><collection>Meteorological & Geoastrophysical Abstracts - Academic</collection><collection>Algology Mycology and Protozoology Abstracts (Microbiology C)</collection><collection>Nursing & Allied Health Premium</collection><collection>Biotechnology and BioEngineering Abstracts</collection><collection>Genetics Abstracts</collection><collection>Environment Abstracts</collection><collection>MEDLINE - Academic</collection><jtitle>Nature (London)</jtitle></facets><delivery><delcategory>Remote Search Resource</delcategory><fulltext>fulltext</fulltext></delivery><addata><au>Wang, Jason Y</au><au>Stevens, Jason M</au><au>Kariofillis, Stavros K</au><au>Tom, Mai-Jan</au><au>Golden, Dung L</au><au>Li, Jun</au><au>Tabora, Jose E</au><au>Parasram, Marvin</au><au>Shields, Benjamin J</au><au>Primer, David N</au><au>Hao, Bo</au><au>Del Valle, David</au><au>DiSomma, Stacey</au><au>Furman, Ariel</au><au>Zipp, G Greg</au><au>Melnikov, Sergey</au><au>Paulson, James</au><au>Doyle, Abigail G</au><format>journal</format><genre>article</genre><ristype>JOUR</ristype><atitle>Identifying general reaction conditions by bandit optimization</atitle><jtitle>Nature (London)</jtitle><addtitle>Nature</addtitle><date>2024-02-29</date><risdate>2024</risdate><volume>626</volume><issue>8001</issue><spage>1025</spage><epage>1033</epage><pages>1025-1033</pages><issn>0028-0836</issn><eissn>1476-4687</eissn><notes>ObjectType-Article-1</notes><notes>SourceType-Scholarly Journals-1</notes><notes>ObjectType-Feature-2</notes><notes>content type line 23</notes><abstract>Reaction conditions that are generally applicable to a wide variety of substrates are highly desired, especially in the pharmaceutical and chemical industries
. Although many approaches are available to evaluate the general applicability of developed conditions, a universal approach to efficiently discover these conditions during optimizations is rare. Here we report the design, implementation and application of reinforcement learning bandit optimization models
to identify generally applicable conditions by efficient condition sampling and evaluation of experimental feedback. Performance benchmarking on existing datasets statistically showed high accuracies for identifying general conditions, with up to 31% improvement over baselines that mimic state-of-the-art optimization approaches. A palladium-catalysed imidazole C-H arylation reaction, an aniline amide coupling reaction and a phenol alkylation reaction were investigated experimentally to evaluate use cases and functionalities of the bandit optimization model in practice. In all three cases, the reaction conditions that were most generally applicable yet not well studied for the respective reaction were identified after surveying less than 15% of the expert-designed reaction space.</abstract><cop>England</cop><pub>Nature Publishing Group</pub><pmid>38418912</pmid><doi>10.1038/s41586-024-07021-y</doi><tpages>9</tpages><orcidid>https://orcid.org/0000-0003-1764-4497</orcidid><orcidid>https://orcid.org/0000-0002-5461-3190</orcidid><orcidid>https://orcid.org/0000-0002-6641-0833</orcidid><orcidid>https://orcid.org/0000-0003-1671-1539</orcidid><orcidid>https://orcid.org/0000-0002-9044-4873</orcidid><orcidid>https://orcid.org/0000-0002-0594-7143</orcidid><orcidid>https://orcid.org/0000-0001-5826-2554</orcidid><orcidid>https://orcid.org/0009-0000-3043-9637</orcidid></addata></record> |
| fulltext | fulltext |
| identifier | ISSN: 0028-0836 |
| ispartof | Nature (London), 2024-02, Vol.626 (8001), p.1025-1033 |
| issn | 0028-0836 1476-4687 |
| language | eng |
| recordid | cdi_proquest_miscellaneous_2933462629 |
| source | Springer Nature - Connect here FIRST to enable access |
| subjects | Algorithms Alkylation Aniline Chemical reactions Chemistry Chemists Datasets Design Gaming machines Imidazole Optimization Optimization models Palladium Phenols Software Substrates |
| title | Identifying general reaction conditions by bandit optimization |
| url | http://sfxeu10.hosted.exlibrisgroup.com/loughborough?ctx_ver=Z39.88-2004&ctx_enc=info:ofi/enc:UTF-8&ctx_tim=2024-09-22T00%3A21%3A39IST&url_ver=Z39.88-2004&url_ctx_fmt=infofi/fmt:kev:mtx:ctx&rfr_id=info:sid/primo.exlibrisgroup.com:primo3-Article-proquest_cross&rft_val_fmt=info:ofi/fmt:kev:mtx:journal&rft.genre=article&rft.atitle=Identifying%20general%20reaction%20conditions%20by%20bandit%20optimization&rft.jtitle=Nature%20(London)&rft.au=Wang,%20Jason%20Y&rft.date=2024-02-29&rft.volume=626&rft.issue=8001&rft.spage=1025&rft.epage=1033&rft.pages=1025-1033&rft.issn=0028-0836&rft.eissn=1476-4687&rft_id=info:doi/10.1038/s41586-024-07021-y&rft_dat=%3Cproquest_cross%3E2937395795%3C/proquest_cross%3E%3Cgrp_id%3Ecdi_FETCH-LOGICAL-c282t-ea04ad522e85a51fb08f51135756ab76cbfda1214df04350cc6006cc6fa8d3ff3%3C/grp_id%3E%3Coa%3E%3C/oa%3E%3Curl%3E%3C/url%3E&rft_id=info:oai/&rft_pqid=2937395795&rft_id=info:pmid/38418912&rfr_iscdi=true |