Automatic classification of experimental models in biomedical literature to support searching for alternative methods to animal experiments

Automatic classification of experimental models in biomedical literature to support searching for alternative methods to animal experiments | Research Square window.SnipcartSettings = { analytics: { enabled: false } }; (function() { var accessVector = localStorage.getItem('access_vector') || ''; window.dataLayer = window.dataLayer || []; if (accessVector) { window.dataLayer.push({ user: { profile: { profileInfo: { snid: accessVector } } } }); } })(); (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start':new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0],j=d.createElement(s),dl=l!='dataLayer'?'&l='+l:'';j.async=true;j.src='https://www.googletagmanager.com/gtm.js?id='+i+dl;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-K279D39R'); Browse Preprints In Review Journals COVID-19 Preprints AJE Video Bytes Research Tools Research Promotion AJE Professional Editing AJE Rubriq About Preprint Platform In Review Editorial Policies Our Team Advisory Board Help Center Sign In Submit a Preprint Cite Share Download PDF Research Article Automatic classification of experimental models in biomedical literature to support searching for alternative methods to animal experiments Mariana Neves, Antonina Klippert, Fanny Knöspel, Juliane Rudeck, and 11 more This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-1526055/v2 This work is licensed under a CC BY 4.0 License Status: Published Journal Publication published 01 Sep, 2023 Read the published version in Journal of Biomedical Semantics → Version 2 posted 7 You are reading this latest preprint version Show more versions Abstract Background: European Union legislature requires replacement of animal experiments with alternative methods, whenever such methods are suitable to reach the intended scientific objective. However, searching for alternative methods in the scientific literature is a time-consuming task that requires careful screening of an enormously large number of experimental biomedical publications. The identification of potentially relevant methods, e.g. organ or cell culture models, or computer simulations, can be supported with text mining tools specifically built for this purpose. Such tools are trained (or fine tuned) on relevant data sets labeled by human experts. Methods: We developed the GoldHamster corpus, composed of 1,600 PubMed (Medline) abstracts, in which we manually identified the used experimental model according to a set of eight labels, namely: "in vivo", "organs", "primary cells", "immortal cell lines", "invertebrates", "humans", "in silico" and "other" (models). We recruited 13 annotators with expertise in the biomedical domain and assigned each article to two individuals. Three additional rounds of annotation aimed at improving the quality of the annotations with disagreements in the first round. Furthermore, we conducted various machine learning experiments based on supervised learning to evaluate the suitability of the corpus for our classification task. Results: We obtained more than 7,000 abstract-level annotations for the above labels. The inter-annotator agreement (kappa coefficient) varied among labels, and ranged from 0.63 (for "others") to 0.82 (for "invertebrates"), with an overall score of 0.74. The best-performing machine learning experiment used the BioBERT pre-trained model with fine-tuning to our corpus, which gained an overall f-score of 0.82. Conclusions: We obtained a high agreement for most of the labels, and our evaluation demonstrated, that our corpus is suitable for training reliable predictive models for automatic classification of biomedical literature according to the used experimental models. Our "Smart feature-based interactive" search tool (SMAFIRA) will employ this classifier for supporting the retrieval of alternative methods to animal experiments. The corpus and the source code will be made available. alternatives to animal experiments corpus annotation text classification replacement Full Text Additional Declarations No competing interests reported. Supplementary Files GoldHamsterSuppMaterial.pdf Cite Share Download PDF Status: Published Journal Publication published 01 Sep, 2023 Read the published version in Journal of Biomedical Semantics → Version 2 posted Editorial decision: Major revision 13 Sep, 2022 Reviews received at journal 24 Aug, 2022 Reviewers agreed at journal 10 Aug, 2022 Reviewers invited by journal 08 Aug, 2022 Submission checks completed at journal 19 Jul, 2022 Editor assigned by journal 19 Jul, 2022 First submitted to journal 15 Jul, 2022 You are reading this latest preprint version Show more versions Research Square lets you share your work early, gain feedback from the community, and start making changes to your manuscript prior to peer review in a journal. As a division of Research Square Company, we’re committed to making research communication faster, fairer, and more useful. We do this by developing innovative software and high quality services for the global research community. Our growing team is made up of researchers and industry professionals working together to solve the most critical problems facing scientific publishing. Also discoverable on Platform About Our Team In Review Editorial Policies Advisory Board Help Center Resources Author Services Accessibility API Access RSS feed Manage Cookie Preferences © Research Square 2026 | ISSN 2693-5015 (online) Privacy Policy Terms of Service Do Not Sell My Personal Information {"props":{"pageProps":{"initialData":{"identity":"rs-1526055","acceptedTermsAndConditions":true,"allowDirectSubmit":false,"archivedVersions":[{"code":1,"date":"2022-04-08 19:11:52","editorialEvents":[{"type":"communityComments","content":0}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"researchsquare","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":true,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"/submission","title":"Research Square","twitterHandle":"researchsquare","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"","reportingPortfolio":"","inReviewEnabled":false,"inReviewRevisionsEnabled":true}}],"articleType":"Research Article","associatedPublications":[],"authors":[{"id":126546240,"identity":"c92dc485-877a-41ab-958d-5861e6d8ad4a","order_by":0,"name":"Mariana Neves","email":"data:image/png;base64,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","orcid":"","institution":"Federal Institute for Risk Assessment","correspondingAuthor":true,"submittingAuthor":false,"prefix":"","firstName":"Mariana","middleName":"","lastName":"Neves","suffix":""},{"id":126546241,"identity":"f55488e4-eff0-4b95-a0a2-494352659f10","order_by":1,"name":"Antonina Klippert","email":"","orcid":"","institution":"Nuvisan ICB GmbH","correspondingAuthor":false,"submittingAuthor":false,"prefix":"","firstName":"Antonina","middleName":"","lastName":"Klippert","suffix":""},{"id":126546242,"identity":"a6facbcb-20eb-494c-a8ae-8c9c4a2f205c","order_by":2,"name":"Fanny Knöspel","email":"","orcid":"","institution":"Federal Institute for Risk Assessment","correspondingAuthor":false,"submittingAuthor":false,"prefix":"","firstName":"Fanny","middleName":"","lastName":"Knöspel","suffix":""},{"id":126546243,"identity":"aa94d3f0-9d40-4021-a211-d4f7f8304183","order_by":3,"name":"Juliane Rudeck","email":"","orcid":"","institution":"Federal Institute for Risk Assessment","correspondingAuthor":false,"submittingAuthor":false,"prefix":"","firstName":"Juliane","middleName":"","lastName":"Rudeck","suffix":""},{"id":126546244,"identity":"b90f4137-b568-485a-823e-57203242dceb","order_by":4,"name":"Ailine Stolz","email":"","orcid":"","institution":"Federal Institute for Risk Assessment","correspondingAuthor":false,"submittingAuthor":false,"prefix":"","firstName":"Ailine","middleName":"","lastName":"Stolz","suffix":""},{"id":126546245,"identity":"667b4a94-7be1-43af-bf29-c130f0e38e7b","order_by":5,"name":"Zsofia Ban","email":"","orcid":"","institution":"Federal Institute for Risk Assessment","correspondingAuthor":false,"submittingAuthor":false,"prefix":"","firstName":"Zsofia","middleName":"","lastName":"Ban","suffix":""},{"id":126546246,"identity":"52c63e6a-bbfa-4b19-9f55-dc45644d56ec","order_by":6,"name":"Markus Becker","email":"","orcid":"","institution":"Federal Institute for Risk Assessment","correspondingAuthor":false,"submittingAuthor":false,"prefix":"","firstName":"Markus","middleName":"","lastName":"Becker","suffix":""},{"id":126546247,"identity":"3ba0c13e-f244-425d-9eba-ccfb89a88ad2","order_by":7,"name":"Kai Diederich","email":"","orcid":"","institution":"Federal Institute for Risk Assessment","correspondingAuthor":false,"submittingAuthor":false,"prefix":"","firstName":"Kai","middleName":"","lastName":"Diederich","suffix":""},{"id":126546248,"identity":"41c69aa5-d7e9-4243-a404-7088e8ccba69","order_by":8,"name":"Barbara Grune","email":"","orcid":"","institution":"Federal Institute for Risk Assessment","correspondingAuthor":false,"submittingAuthor":false,"prefix":"","firstName":"Barbara","middleName":"","lastName":"Grune","suffix":""},{"id":126546249,"identity":"50dd7006-855d-4edd-9eb2-569dd30fbd07","order_by":9,"name":"Pia Kahnau","email":"","orcid":"","institution":"Federal Institute for Risk Assessment","correspondingAuthor":false,"submittingAuthor":false,"prefix":"","firstName":"Pia","middleName":"","lastName":"Kahnau","suffix":""},{"id":126546250,"identity":"ee2b8326-2a6d-418d-86f9-c1f0df30ad4c","order_by":10,"name":"Nils Ohnesorge","email":"","orcid":"","institution":"Federal Institute for Risk Assessment","correspondingAuthor":false,"submittingAuthor":false,"prefix":"","firstName":"Nils","middleName":"","lastName":"Ohnesorge","suffix":""},{"id":126546251,"identity":"6a3bdcef-2644-4f9a-8eaa-cd8866fff675","order_by":11,"name":"Johannes Pucher","email":"","orcid":"","institution":"Federal Institute for Risk Assessment","correspondingAuthor":false,"submittingAuthor":false,"prefix":"","firstName":"Johannes","middleName":"","lastName":"Pucher","suffix":""},{"id":126546252,"identity":"7badfe3d-801b-4e1e-b0e8-aab28b76bb70","order_by":12,"name":"Gilbert Schönfelder","email":"","orcid":"","institution":"Federal Institute for Risk Assessment","correspondingAuthor":false,"submittingAuthor":false,"prefix":"","firstName":"Gilbert","middleName":"","lastName":"Schönfelder","suffix":""},{"id":126546253,"identity":"e1f7461e-62f3-4c45-ac19-6a0ce7a47bf3","order_by":13,"name":"Bettina Bert","email":"","orcid":"","institution":"Federal Institute for Risk Assessment","correspondingAuthor":false,"submittingAuthor":false,"prefix":"","firstName":"Bettina","middleName":"","lastName":"Bert","suffix":""},{"id":126546254,"identity":"8553e21d-f650-41b2-880c-54377f266189","order_by":14,"name":"Daniel Butzke","email":"","orcid":"","institution":"Federal Institute for Risk Assessment","correspondingAuthor":false,"submittingAuthor":false,"prefix":"","firstName":"Daniel","middleName":"","lastName":"Butzke","suffix":""}],"badges":[],"createdAt":"2022-04-05 13:59:22","currentVersionCode":2,"declarations":"","doi":"10.21203/rs.3.rs-1526055/v2","doiUrl":"https://doi.org/10.21203/rs.3.rs-1526055/v2","draftVersion":[],"editorialEvents":[{"content":"https://doi.org/10.1186/s13326-023-00292-w","type":"published","date":"2023-09-01T15:10:14+00:00"}],"editorialNote":"","failedWorkflow":false,"files":[{"id":24856980,"identity":"841dcb62-b014-4d9a-afa1-feea0bf95d8d","added_by":"auto","created_at":"2022-08-05 23:35:44","extension":"pdf","order_by":1,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":697108,"visible":true,"origin":"","legend":"","description":"","filename":"GoldHamsterJBMS.pdf","url":"https://assets-eu.researchsquare.com/files/rs-1526055/v2_covered.pdf"},{"id":24856979,"identity":"31bf44a1-fc6d-435c-a744-0d196448d1bd","added_by":"auto","created_at":"2022-08-05 23:35:39","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"supplement","size":798571,"visible":true,"origin":"","legend":"","description":"","filename":"GoldHamsterSuppMaterial.pdf","url":"https://assets-eu.researchsquare.com/files/rs-1526055/v2/e3e0a578da51d3b032abd57d.pdf"}],"financialInterests":"No competing interests reported.","formattedTitle":"Automatic classification of experimental models in biomedical literature to support searching for alternative methods to animal experiments","fulltext":[{"header":"Full Text","content":"This preprint is available for \u003ca href='/article/rs-1526055/latest.pdf' target='_blank'\u003edownload as a PDF\u003c/a\u003e."}],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":false,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":false,"hideJournal":false,"highlight":"","institution":"","isAcceptedByJournal":true,"isAuthorSuppliedPdf":true,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":false,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"journal-of-biomedical-semantics","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"jbsm","sideBox":"Learn more about [Journal of Biomedical Semantics](http://jbiomedsem.biomedcentral.com/)","snPcode":"13326","submissionUrl":"https://submission.nature.com/new-submission/13326/3","title":"Journal of Biomedical Semantics","twitterHandle":"","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"em","reportingPortfolio":"BMC/SO AJ","inReviewEnabled":true,"inReviewRevisionsEnabled":true},"keywords":"alternatives to animal experiments, corpus annotation, text classification, replacement","lastPublishedDoi":"10.21203/rs.3.rs-1526055/v2","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-1526055/v2","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003cp\u003e\u003cstrong\u003eBackground: \u003c/strong\u003eEuropean Union legislature requires replacement of animal experiments with alternative methods, whenever such methods are suitable to reach the intended scientific objective. However, searching for alternative methods in the scientific literature is a time-consuming task that requires careful screening of an enormously large number of experimental biomedical publications. The identification of potentially relevant methods, e.g. organ or cell culture models, or computer simulations, can be supported with text mining tools specifically built for this purpose. Such tools are trained (or fine tuned) on relevant data sets labeled by human experts. \u003c/p\u003e\u003cp\u003e\u003cstrong\u003eMethods: \u003c/strong\u003eWe developed the GoldHamster corpus, composed of 1,600 PubMed (Medline) abstracts, in which we manually identified the used experimental model according to a set of eight labels, namely: \"in vivo\", \"organs\", \"primary cells\", \"immortal cell lines\", \"invertebrates\", \"humans\", \"in silico\" and \"other\" (models). We recruited 13 annotators with expertise in the biomedical domain and assigned each article to two individuals. Three additional rounds of annotation aimed at improving the quality of the annotations with disagreements in the first round. Furthermore, we conducted various machine learning experiments based on supervised learning to evaluate the suitability of the corpus for our classification task. \u003c/p\u003e\u003cp\u003e\u003cstrong\u003eResults: \u003c/strong\u003eWe obtained more than 7,000 abstract-level annotations for the above labels. The inter-annotator agreement (kappa coefficient) varied among labels, and ranged from 0.63 (for \"others\") to 0.82 (for \"invertebrates\"), with an overall score of 0.74. The best-performing machine learning experiment used the BioBERT pre-trained model with fine-tuning to our corpus, which gained an overall f-score of 0.82. \u003c/p\u003e\u003cp\u003e\u003cstrong\u003eConclusions: \u003c/strong\u003eWe obtained a high agreement for most of the labels, and our evaluation demonstrated, that our corpus is suitable for training reliable predictive models for automatic classification of biomedical literature according to the used experimental models. Our \"Smart feature-based interactive\" search tool (SMAFIRA) will employ this classifier for supporting the retrieval of alternative methods to animal experiments. The corpus and the source code will be made available.\u003c/p\u003e","manuscriptTitle":"Automatic classification of experimental models in biomedical literature to support searching for alternative methods to animal experiments","msid":"","msnumber":"","nonDraftVersions":[{"code":2,"date":"2022-08-05 23:35:37","doi":"10.21203/rs.3.rs-1526055/v2","editorialEvents":[{"type":"communityComments","content":0},{"type":"decision","content":"Major revision","date":"2022-09-13T15:14:54+00:00","index":"","fulltext":""},{"type":"editorInvitedReview","content":"","date":"2022-08-24T20:31:28+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"67fd897c-24a6-4b2e-abda-5526222c9484","date":"2022-08-10T04:36:36+00:00","index":"hide","fulltext":""},{"type":"reviewersInvited","content":"","date":"2022-08-08T14:41:19+00:00","index":"","fulltext":""},{"type":"checksComplete","content":"","date":"2022-07-20T01:46:55+00:00","index":"","fulltext":""},{"type":"editorAssigned","content":"","date":"2022-07-20T01:46:55+00:00","index":"","fulltext":""},{"type":"submitted","content":"Journal of Biomedical Semantics","date":"2022-07-15T09:27:53+00:00","index":"","fulltext":""}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"journal-of-biomedical-semantics","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"jbsm","sideBox":"Learn more about [Journal of Biomedical Semantics](http://jbiomedsem.biomedcentral.com/)","snPcode":"13326","submissionUrl":"https://submission.nature.com/new-submission/13326/3","title":"Journal of Biomedical Semantics","twitterHandle":"","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"em","reportingPortfolio":"BMC/SO AJ","inReviewEnabled":true,"inReviewRevisionsEnabled":true}}],"origin":"","ownerIdentity":"d744b6c3-6f7f-4c20-979c-1e68fcee1b8d","owner":[],"postedDate":"August 5th, 2022","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"published-in-journal","subjectAreas":[],"tags":[],"updatedAt":"2023-09-07T15:15:40+00:00","versionOfRecord":{"articleIdentity":"rs-1526055","link":"https://doi.org/10.1186/s13326-023-00292-w","journal":{"identity":"journal-of-biomedical-semantics","isVorOnly":false,"title":"Journal of Biomedical Semantics"},"publishedOn":"2023-09-01 15:10:14","publishedOnDateReadable":"September 1st, 2023"},"versionCreatedAt":"2022-08-05 23:35:37","video":"","vorDoi":"10.1186/s13326-023-00292-w","vorDoiUrl":"https://doi.org/10.1186/s13326-023-00292-w","workflowStages":[]},"version":"v2","identity":"rs-1526055","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-1526055","identity":"rs-1526055","version":["v2"]},"buildId":"-HB7Z8yhvgn0wM9Nzuekk","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}