{"id":"https://openalex.org/W4410003735","doi":"https://doi.org/10.1021/acs.jcim.5c00359","title":"Exploring BERT for Reaction Yield Prediction: Evaluating the Impact of Tokenization, Molecular Representation, and Pretraining Data Augmentation","display_name":"Exploring BERT for Reaction Yield Prediction: Evaluating the Impact of Tokenization, Molecular Representation, and Pretraining Data Augmentation","publication_year":2025,"publication_date":"2025-05-01","ids":{"openalex":"https://openalex.org/W4410003735","doi":"https://doi.org/10.1021/acs.jcim.5c00359","pmid":"https://pubmed.ncbi.nlm.nih.gov/40311104"},"language":"en","primary_location":{"id":"doi:10.1021/acs.jcim.5c00359","is_oa":false,"landing_page_url":"https://doi.org/10.1021/acs.jcim.5c00359","pdf_url":null,"source":{"id":"https://openalex.org/S167262187","display_name":"Journal of Chemical Information and Modeling","issn_l":"1549-9596","issn":["1549-9596","1549-960X"],"is_oa":false,"is_in_doaj":false,"is_core":true,"host_organization":"https://openalex.org/P4310320006","host_organization_name":"American Chemical Society","host_organization_lineage":["https://openalex.org/P4310320006"],"host_organization_lineage_names":["American Chemical Society"],"type":"journal"},"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"Journal of Chemical Information and Modeling","raw_type":"journal-article"},"type":"article","indexed_in":["crossref","pubmed"],"open_access":{"is_oa":false,"oa_status":"closed","oa_url":null,"any_repository_has_fulltext":false},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5076430061","display_name":"Adrian Krzyzanowski","orcid":"https://orcid.org/0000-0002-3604-9274"},"institutions":[{"id":"https://openalex.org/I2800110054","display_name":"Age UK","ror":"https://ror.org/050x9d346","country_code":"GB","type":"nonprofit","lineage":["https://openalex.org/I2800110054"]}],"countries":["GB"],"is_corresponding":true,"raw_author_name":"Adrian Krzyzanowski","raw_affiliation_strings":["GSK Medicines Research Centre, Gunnels Wood Road, Stevenage SG1 2NY, U.K"],"affiliations":[{"raw_affiliation_string":"GSK Medicines Research Centre, Gunnels Wood Road, Stevenage SG1 2NY, U.K","institution_ids":["https://openalex.org/I2800110054"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5079075074","display_name":"Stephen D. Pickett","orcid":"https://orcid.org/0000-0002-0958-9830"},"institutions":[{"id":"https://openalex.org/I2800110054","display_name":"Age UK","ror":"https://ror.org/050x9d346","country_code":"GB","type":"nonprofit","lineage":["https://openalex.org/I2800110054"]}],"countries":["GB"],"is_corresponding":false,"raw_author_name":"Stephen D. Pickett","raw_affiliation_strings":["GSK Medicines Research Centre, Gunnels Wood Road, Stevenage SG1 2NY, U.K"],"affiliations":[{"raw_affiliation_string":"GSK Medicines Research Centre, Gunnels Wood Road, Stevenage SG1 2NY, U.K","institution_ids":["https://openalex.org/I2800110054"]}]},{"author_position":"last","author":{"id":"https://openalex.org/A5032112911","display_name":"P\u00e9ter Pog\u00e1ny","orcid":"https://orcid.org/0000-0003-3536-0746"},"institutions":[{"id":"https://openalex.org/I2800110054","display_name":"Age UK","ror":"https://ror.org/050x9d346","country_code":"GB","type":"nonprofit","lineage":["https://openalex.org/I2800110054"]}],"countries":["GB"],"is_corresponding":true,"raw_author_name":"Peter Pog\u00e1ny","raw_affiliation_strings":["GSK Medicines Research Centre, Gunnels Wood Road, Stevenage SG1 2NY, U.K"],"affiliations":[{"raw_affiliation_string":"GSK Medicines Research Centre, Gunnels Wood Road, Stevenage SG1 2NY, U.K","institution_ids":["https://openalex.org/I2800110054"]}]}],"institutions":[],"countries_distinct_count":1,"institutions_distinct_count":3,"corresponding_author_ids":["https://openalex.org/A5032112911","https://openalex.org/A5076430061"],"corresponding_institution_ids":["https://openalex.org/I2800110054"],"apc_list":null,"apc_paid":null,"fwci":2.7026,"has_fulltext":false,"cited_by_count":5,"citation_normalized_percentile":{"value":0.9006988,"is_in_top_1_percent":false,"is_in_top_10_percent":true},"cited_by_percentile_year":{"min":95,"max":100},"biblio":{"volume":"65","issue":"9","first_page":"4381","last_page":"4402"},"is_retracted":false,"is_paratext":false,"is_xpac":false,"primary_topic":{"id":"https://openalex.org/T11948","display_name":"Machine Learning in Materials Science","score":0.9988999962806702,"subfield":{"id":"https://openalex.org/subfields/2505","display_name":"Materials Chemistry"},"field":{"id":"https://openalex.org/fields/25","display_name":"Materials Science"},"domain":{"id":"https://openalex.org/domains/3","display_name":"Physical Sciences"}},"topics":[{"id":"https://openalex.org/T11948","display_name":"Machine Learning in Materials Science","score":0.9988999962806702,"subfield":{"id":"https://openalex.org/subfields/2505","display_name":"Materials Chemistry"},"field":{"id":"https://openalex.org/fields/25","display_name":"Materials Science"},"domain":{"id":"https://openalex.org/domains/3","display_name":"Physical Sciences"}},{"id":"https://openalex.org/T10211","display_name":"Computational Drug Discovery Methods","score":0.9984999895095825,"subfield":{"id":"https://openalex.org/subfields/1703","display_name":"Computational Theory and Mathematics"},"field":{"id":"https://openalex.org/fields/17","display_name":"Computer Science"},"domain":{"id":"https://openalex.org/domains/3","display_name":"Physical Sciences"}},{"id":"https://openalex.org/T10908","display_name":"Analytical Chemistry and Chromatography","score":0.9294999837875366,"subfield":{"id":"https://openalex.org/subfields/1607","display_name":"Spectroscopy"},"field":{"id":"https://openalex.org/fields/16","display_name":"Chemistry"},"domain":{"id":"https://openalex.org/domains/3","display_name":"Physical Sciences"}}],"keywords":[{"id":"https://openalex.org/keywords/lexical-analysis","display_name":"Lexical analysis","score":0.7467406392097473},{"id":"https://openalex.org/keywords/yield","display_name":"Yield (engineering)","score":0.6695098280906677},{"id":"https://openalex.org/keywords/computer-science","display_name":"Computer science","score":0.616071343421936},{"id":"https://openalex.org/keywords/representation","display_name":"Representation (politics)","score":0.6129463315010071},{"id":"https://openalex.org/keywords/artificial-intelligence","display_name":"Artificial intelligence","score":0.47446805238723755},{"id":"https://openalex.org/keywords/natural-language-processing","display_name":"Natural language processing","score":0.4131756126880646},{"id":"https://openalex.org/keywords/machine-learning","display_name":"Machine learning","score":0.34285059571266174},{"id":"https://openalex.org/keywords/materials-science","display_name":"Materials science","score":0.09656748175621033}],"concepts":[{"id":"https://openalex.org/C176982825","wikidata":"https://www.wikidata.org/wiki/Q835922","display_name":"Lexical analysis","level":2,"score":0.7467406392097473},{"id":"https://openalex.org/C134121241","wikidata":"https://www.wikidata.org/wiki/Q899301","display_name":"Yield (engineering)","level":2,"score":0.6695098280906677},{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.616071343421936},{"id":"https://openalex.org/C2776359362","wikidata":"https://www.wikidata.org/wiki/Q2145286","display_name":"Representation (politics)","level":3,"score":0.6129463315010071},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.47446805238723755},{"id":"https://openalex.org/C204321447","wikidata":"https://www.wikidata.org/wiki/Q30642","display_name":"Natural language processing","level":1,"score":0.4131756126880646},{"id":"https://openalex.org/C119857082","wikidata":"https://www.wikidata.org/wiki/Q2539","display_name":"Machine learning","level":1,"score":0.34285059571266174},{"id":"https://openalex.org/C192562407","wikidata":"https://www.wikidata.org/wiki/Q228736","display_name":"Materials science","level":0,"score":0.09656748175621033},{"id":"https://openalex.org/C94625758","wikidata":"https://www.wikidata.org/wiki/Q7163","display_name":"Politics","level":2,"score":0.0},{"id":"https://openalex.org/C191897082","wikidata":"https://www.wikidata.org/wiki/Q11467","display_name":"Metallurgy","level":1,"score":0.0},{"id":"https://openalex.org/C199539241","wikidata":"https://www.wikidata.org/wiki/Q7748","display_name":"Law","level":1,"score":0.0},{"id":"https://openalex.org/C17744445","wikidata":"https://www.wikidata.org/wiki/Q36442","display_name":"Political science","level":0,"score":0.0}],"mesh":[{"descriptor_ui":"D000069550","descriptor_name":"Machine Learning","qualifier_ui":null,"qualifier_name":null,"is_major_topic":false},{"descriptor_ui":"D000069550","descriptor_name":"Machine Learning","qualifier_ui":null,"qualifier_name":null,"is_major_topic":false},{"descriptor_ui":"D060326","descriptor_name":"Chemistry Techniques, Synthetic","qualifier_ui":"Q000379","qualifier_name":"methods","is_major_topic":true},{"descriptor_ui":"D060326","descriptor_name":"Chemistry Techniques, Synthetic","qualifier_ui":"Q000379","qualifier_name":"methods","is_major_topic":true}],"locations_count":2,"locations":[{"id":"doi:10.1021/acs.jcim.5c00359","is_oa":false,"landing_page_url":"https://doi.org/10.1021/acs.jcim.5c00359","pdf_url":null,"source":{"id":"https://openalex.org/S167262187","display_name":"Journal of Chemical Information and Modeling","issn_l":"1549-9596","issn":["1549-9596","1549-960X"],"is_oa":false,"is_in_doaj":false,"is_core":true,"host_organization":"https://openalex.org/P4310320006","host_organization_name":"American Chemical Society","host_organization_lineage":["https://openalex.org/P4310320006"],"host_organization_lineage_names":["American Chemical Society"],"type":"journal"},"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"Journal of Chemical Information and Modeling","raw_type":"journal-article"},{"id":"pmid:40311104","is_oa":false,"landing_page_url":"https://pubmed.ncbi.nlm.nih.gov/40311104","pdf_url":null,"source":{"id":"https://openalex.org/S4306525036","display_name":"PubMed","issn_l":null,"issn":null,"is_oa":false,"is_in_doaj":false,"is_core":false,"host_organization":"https://openalex.org/I1299303238","host_organization_name":"National Institutes of Health","host_organization_lineage":["https://openalex.org/I1299303238"],"host_organization_lineage_names":[],"type":"repository"},"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"Journal of chemical information and modeling","raw_type":null}],"best_oa_location":null,"sustainable_development_goals":[],"awards":[],"funders":[{"id":"https://openalex.org/F4320307773","display_name":"GlaxoSmithKline","ror":"https://ror.org/01xsqw823"}],"has_content":{"grobid_xml":false,"pdf":false},"content_urls":null,"referenced_works_count":77,"referenced_works":["https://openalex.org/W29374554","https://openalex.org/W58593274","https://openalex.org/W103650626","https://openalex.org/W1757990252","https://openalex.org/W1974758710","https://openalex.org/W1975147762","https://openalex.org/W1980361506","https://openalex.org/W1988037271","https://openalex.org/W2016944307","https://openalex.org/W2121879602","https://openalex.org/W2155464463","https://openalex.org/W2186518338","https://openalex.org/W2295598076","https://openalex.org/W2525778437","https://openalex.org/W2622322262","https://openalex.org/W2750779823","https://openalex.org/W2769423117","https://openalex.org/W2784918212","https://openalex.org/W2785942661","https://openalex.org/W2895677720","https://openalex.org/W2900743800","https://openalex.org/W2911489562","https://openalex.org/W2924700250","https://openalex.org/W2937845937","https://openalex.org/W2947423323","https://openalex.org/W2965373594","https://openalex.org/W2979826702","https://openalex.org/W2989615256","https://openalex.org/W2996428491","https://openalex.org/W3003243900","https://openalex.org/W3009321976","https://openalex.org/W3012519883","https://openalex.org/W3017003177","https://openalex.org/W3038678525","https://openalex.org/W3045928028","https://openalex.org/W3046375318","https://openalex.org/W3091739897","https://openalex.org/W3093934881","https://openalex.org/W3097598035","https://openalex.org/W3103092523","https://openalex.org/W3118349318","https://openalex.org/W3118543624","https://openalex.org/W3143418323","https://openalex.org/W3143644407","https://openalex.org/W3199705903","https://openalex.org/W4206063340","https://openalex.org/W4220699355","https://openalex.org/W4220902634","https://openalex.org/W4220918529","https://openalex.org/W4225764472","https://openalex.org/W4226145240","https://openalex.org/W4229044212","https://openalex.org/W4239019441","https://openalex.org/W4293580221","https://openalex.org/W4297814571","https://openalex.org/W4300996741","https://openalex.org/W4311728218","https://openalex.org/W4315645746","https://openalex.org/W4320036219","https://openalex.org/W4324122028","https://openalex.org/W4362664882","https://openalex.org/W4380870591","https://openalex.org/W4382808681","https://openalex.org/W4386316087","https://openalex.org/W4386316902","https://openalex.org/W4386702940","https://openalex.org/W4389472086","https://openalex.org/W4389996830","https://openalex.org/W4390501063","https://openalex.org/W4390959384","https://openalex.org/W4391160649","https://openalex.org/W4393317648","https://openalex.org/W4394485229","https://openalex.org/W4398145691","https://openalex.org/W4401689901","https://openalex.org/W4403138423","https://openalex.org/W4405172829"],"related_works":["https://openalex.org/W4405003489","https://openalex.org/W4386014872","https://openalex.org/W1847536016","https://openalex.org/W4361193986","https://openalex.org/W4300598845","https://openalex.org/W2601638452","https://openalex.org/W2285263069","https://openalex.org/W4376107815","https://openalex.org/W4319309671","https://openalex.org/W4319309603"],"abstract_inverted_index":{"Predicting":[0],"reaction":[1,125,129,175],"yields":[2],"in":[3,194],"synthetic":[4,195],"chemistry":[5],"remains":[6],"a":[7,26,120,158],"significant":[8],"challenge.":[9],"This":[10],"study":[11],"systematically":[12],"evaluates":[13],"the":[14,124,147,150,206],"impact":[15,60],"of":[16,32,101,105],"tokenization,":[17],"molecular":[18,47],"representation,":[19],"pretraining":[20,83,109,144],"data,":[21],"and":[22,34,67,79,149,170,174,186],"adversarial":[23,160],"training":[24,161,199],"on":[25,61],"BERT-based":[27],"model":[28,62,168],"for":[29,77,123,142,172,183,191],"yield":[30,173,192],"prediction":[31,193],"Buchwald-Hartwig":[33],"Suzuki-Miyaura":[35],"coupling":[36],"reactions":[37],"using":[38],"publicly":[39],"available":[40,204],"HTE":[41],"data":[42,87,97,110,130],"sets.":[43],"We":[44],"demonstrate":[45],"that":[46,157],"representation":[48],"choice":[49],"(SMILES,":[50],"DeepSMILES,":[51],"SELFIES,":[52],"Morgan":[53],"fingerprint-based":[54,80],"notation,":[55],"IUPAC":[56],"names)":[57],"has":[58],"minimal":[59],"performance,":[63],"while":[64],"typically":[65],"BPE":[66],"SentencePiece":[68],"tokenization":[69],"outperform":[70],"other":[71],"methods.":[72],"WordPiece":[73],"is":[74,111,202],"strongly":[75],"discouraged":[76],"SELFIES":[78],"notation.":[81],"Furthermore,":[82],"with":[84,208],"relatively":[85],"small":[86],"sets":[88,98,116,131,145],"(<100":[89],"K":[90],"reactions)":[91],"achieves":[92],"comparable":[93],"performance":[94,139],"to":[95,118,205],"larger":[96],"containing":[99],"millions":[100],"examples.":[102],"The":[103,113,137],"use":[104],"artificially":[106,114],"generated":[107,115],"domain-specific":[108],"proposed.":[112],"prove":[117],"be":[119],"good":[121],"surrogate":[122],"schemes":[126],"extracted":[127],"from":[128],"such":[132],"as":[133],"Pistachio":[134],"or":[135],"Reaxys.":[136],"best":[138],"was":[140],"observed":[141],"hybrid":[143],"combining":[146],"real":[148],"domain-specific,":[151],"artificial":[152],"data.":[153],"Finally,":[154],"we":[155],"show":[156],"novel":[159],"approach,":[162],"perturbing":[163],"input":[164],"embeddings":[165],"dynamically,":[166],"improves":[167],"robustness":[169],"generalizability":[171],"success":[176],"prediction.":[177],"These":[178],"findings":[179],"provide":[180],"valuable":[181],"insights":[182],"developing":[184],"robust":[185],"practical":[187],"machine":[188],"learning":[189],"models":[190],"chemistry.":[196],"GSK's":[197],"BERT":[198],"code":[200],"base":[201],"made":[203],"community":[207],"this":[209],"work.":[210]},"counts_by_year":[{"year":2026,"cited_by_count":3},{"year":2025,"cited_by_count":2}],"updated_date":"2026-02-25T08:12:03.925757","created_date":"2025-10-10T00:00:00"}