Model-guided optimization of a low-serum medium for Madin–Darby canine kidney cells using a Kolmogorov–Arnold network and Bayesian optimization

Model-guided optimization of a low-serum medium for Madin–Darby canine kidney cells using a Kolmogorov–Arnold network and Bayesian optimization | 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 Model-guided optimization of a low-serum medium for Madin–Darby canine kidney cells using a Kolmogorov–Arnold network and Bayesian optimization Mengting Zhang, Juntao Wang, Lei Zhou, Jianmin Chen, Qilei Liu, and 1 more This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-8549568/v1 This work is licensed under a CC BY 4.0 License Status: Under Review Version 1 posted 5 You are reading this latest preprint version Abstract Madin–Darby canine kidney (MDCK) cells are usually grown in a medium containing a high concentration of foetal bovine serum (FBS). However, this can lead to problems such as higher costs, inconsistent results between batches, and regulatory issues. This study presents a framework for developing and optimising a low-serum medium that combines systematic experimentation with machine learning. Initial screening of 27 nutritional components using a Plackett–Burman design revealed that glutamine, tyrosine and asparagine were critical factors. A Box–Behnken design was then employed to generate training data, modelled using a Kolmogorov–Arnold network (KAN), and optimized via Bayesian optimization (BO) to efficiently navigate the parameter space. The final formulation was found to significantly promote cell proliferation, achieving increases of 58.6% and 7.2% compared to in-house and commercial 10% FBS baselines, respectively. In virus production assays, the optimized medium yielded peak BVR-26 titres of 3.52 ± 0.06 log₁₀ TCID₅₀/mL at 48 hours post-infection, outperforming the commercial control. Metabolic profiling revealed enhanced glucose utilisation and reduced lactate accumulation, corroborating the formulation's physiological benefits. This work establishes a robust, transferable platform for accelerating the development of low-serum bioprocess media by effectively integrating machine learning with structured experimental design. MDCK cells low-serum medium design of experiments Bayesian optimisation Kolmogorov–Arnold network viral titer Full Text Cite Share Download PDF Status: Under Review Version 1 posted Reviewers agreed at journal 12 May, 2026 Reviewers invited by journal 01 Mar, 2026 Editor invited by journal 14 Jan, 2026 Editor assigned by journal 14 Jan, 2026 First submitted to journal 10 Jan, 2026 You are reading this latest preprint version 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-8549568","acceptedTermsAndConditions":true,"allowDirectSubmit":false,"archivedVersions":[],"articleType":"Research Article","associatedPublications":[],"authors":[{"id":598998447,"identity":"2bfb2571-42af-4d21-bfa5-69702b335335","order_by":0,"name":"Mengting Zhang","email":"","orcid":"","institution":"Dalian University of Technology","correspondingAuthor":false,"prefix":"","firstName":"Mengting","middleName":"","lastName":"Zhang","suffix":""},{"id":598998448,"identity":"e1f650e1-4753-414d-855c-4ab7af074b91","order_by":1,"name":"Juntao Wang","email":"","orcid":"","institution":"Dalian University of Technology","correspondingAuthor":false,"prefix":"","firstName":"Juntao","middleName":"","lastName":"Wang","suffix":""},{"id":598998449,"identity":"244bd5b6-afdf-486f-8542-00bc055ef46d","order_by":2,"name":"Lei Zhou","email":"","orcid":"","institution":"Dalian University of Technology","correspondingAuthor":false,"prefix":"","firstName":"Lei","middleName":"","lastName":"Zhou","suffix":""},{"id":598998450,"identity":"c4b1dc8d-2b54-4ac1-a2ed-6b126b615a45","order_by":3,"name":"Jianmin Chen","email":"","orcid":"","institution":"Fosun Apexvac (Dalian）Biopharmaceutical Co.,Ltd.","correspondingAuthor":false,"prefix":"","firstName":"Jianmin","middleName":"","lastName":"Chen","suffix":""},{"id":598998451,"identity":"d68fa122-5403-4f6d-b7d1-f903e54893eb","order_by":4,"name":"Qilei Liu","email":"","orcid":"","institution":"Dalian University of Technology","correspondingAuthor":false,"prefix":"","firstName":"Qilei","middleName":"","lastName":"Liu","suffix":""},{"id":598998452,"identity":"5ee0029a-cef9-4a3e-8c80-c05bb3dbe396","order_by":5,"name":"Qingwei Meng","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAAAx0lEQVRIiWNgGAWjYPACG8Y+MM1GvJY0xjYGZtK0HCZBi2577+EXb9vOy7ZJ5B9g+FB2mIF/dgN+LWZnzqVZzm27bdwmkczAOOPcYQaJOwcIaLmRY2bMu+12IkgLM2/bYQYDiQQCWu6/AWk5B9HylygtN3iMH/NuOwDRwkiUljM5Zoxz/yUbt/E8NjjYcy6dR+IGIS3Hzxh/eHPGTrafPfHhgx9l1nL8MwhoAQI2CR4o6wAQ8+BRCQfMH4hSNgpGwSgYBSMXAACi+kMUwOmfMwAAAABJRU5ErkJggg==","orcid":"","institution":"Dalian University of Technology","correspondingAuthor":true,"prefix":"","firstName":"Qingwei","middleName":"","lastName":"Meng","suffix":""}],"badges":[],"createdAt":"2026-01-08 09:26:25","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-8549568/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-8549568/v1","draftVersion":[],"editorialEvents":[],"editorialNote":"","failedWorkflow":false,"files":[{"id":103909429,"identity":"b8a95db8-3557-42f3-998a-80539fb47aac","added_by":"auto","created_at":"2026-03-04 11:43:05","extension":"pdf","order_by":1,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":897217,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-8549568/v1_covered_a8763594-b9b4-4504-81b3-68c2fb390954.pdf"}],"financialInterests":"","formattedTitle":"Model-guided optimization of a low-serum medium for Madin–Darby canine kidney cells using a Kolmogorov–Arnold network and Bayesian optimization","fulltext":[],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":false,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":false,"hideJournal":false,"highlight":"","institution":"","isAcceptedByJournal":false,"isAuthorSuppliedPdf":true,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":true,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"applied-biochemistry-and-biotechnology","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"abab","sideBox":"Learn more about [Applied Biochemistry and Biotechnology](https://www.springer.com/journal/12010)","snPcode":"12010","submissionUrl":"https://submission.nature.com/new-submission/12010/3","title":"Applied Biochemistry and Biotechnology","twitterHandle":"","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"em","reportingPortfolio":"Springer Hybrid","inReviewEnabled":true,"inReviewRevisionsEnabled":false},"keywords":"MDCK cells, low-serum medium, design of experiments, Bayesian optimisation, Kolmogorov–Arnold network, viral titer","lastPublishedDoi":"10.21203/rs.3.rs-8549568/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-8549568/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003cp\u003eMadin\u0026ndash;Darby canine kidney (MDCK) cells are usually grown in a medium containing a high concentration of foetal bovine serum (FBS). However, this can lead to problems such as higher costs, inconsistent results between batches, and regulatory issues. This study presents a framework for developing and optimising a low-serum medium that combines systematic experimentation with machine learning. Initial screening of 27 nutritional components using a Plackett\u0026ndash;Burman design revealed that glutamine, tyrosine and asparagine were critical factors. A Box\u0026ndash;Behnken design was then employed to generate training data, modelled using a Kolmogorov\u0026ndash;Arnold network (KAN), and optimized via Bayesian optimization (BO) to efficiently navigate the parameter space. The final formulation was found to significantly promote cell proliferation, achieving increases of 58.6% and 7.2% compared to in-house and commercial 10% FBS baselines, respectively. In virus production assays, the optimized medium yielded peak BVR-26 titres of 3.52\u0026thinsp;\u0026plusmn;\u0026thinsp;0.06 log₁₀ TCID₅₀/mL at 48 hours post-infection, outperforming the commercial control. Metabolic profiling revealed enhanced glucose utilisation and reduced lactate accumulation, corroborating the formulation's physiological benefits. This work establishes a robust, transferable platform for accelerating the development of low-serum bioprocess media by effectively integrating machine learning with structured experimental design.\u003c/p\u003e","manuscriptTitle":"Model-guided optimization of a low-serum medium for Madin–Darby canine kidney cells using a Kolmogorov–Arnold network and Bayesian optimization","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2026-03-04 11:41:54","doi":"10.21203/rs.3.rs-8549568/v1","editorialEvents":[{"type":"communityComments","content":0},{"type":"reviewerAgreed","content":"","date":"2026-05-12T14:10:45+00:00","index":0,"fulltext":""},{"type":"reviewersInvited","content":"","date":"2026-03-02T04:41:53+00:00","index":"","fulltext":""},{"type":"editorInvited","content":"Applied Biochemistry and Biotechnology","date":"2026-01-14T15:55:42+00:00","index":"","fulltext":""},{"type":"editorAssigned","content":"","date":"2026-01-14T05:05:14+00:00","index":"","fulltext":""},{"type":"submitted","content":"Applied Biochemistry and Biotechnology","date":"2026-01-11T00:40:32+00:00","index":"","fulltext":""}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"applied-biochemistry-and-biotechnology","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"abab","sideBox":"Learn more about [Applied Biochemistry and Biotechnology](https://www.springer.com/journal/12010)","snPcode":"12010","submissionUrl":"https://submission.nature.com/new-submission/12010/3","title":"Applied Biochemistry and Biotechnology","twitterHandle":"","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"em","reportingPortfolio":"Springer Hybrid","inReviewEnabled":true,"inReviewRevisionsEnabled":false}}],"origin":"","ownerIdentity":"0323c944-f968-4d55-8495-b37dd4e913f9","owner":[],"postedDate":"March 4th, 2026","published":true,"recentEditorialEvents":[{"type":"reviewerAgreed","content":"","date":"2026-05-12T14:10:45+00:00","index":0,"fulltext":""}],"rejectedJournal":[],"revision":"","amendment":"","status":"under-review","subjectAreas":[],"tags":[],"updatedAt":"2026-03-04T11:41:55+00:00","versionOfRecord":[],"versionCreatedAt":"2026-03-04 11:41:54","video":"","vorDoi":"","vorDoiUrl":"","workflowStages":[]},"version":"v1","identity":"rs-8549568","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-8549568","identity":"rs-8549568","version":["v1"]},"buildId":"XKTyCvWXoU3ODBz1xrDgd","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}