A Deep Learning Approach to Quantitative PCR that Learns from Ground Truth

A Deep Learning Approach to Quantitative PCR that Learns from Ground Truth | 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 Article A Deep Learning Approach to Quantitative PCR that Learns from Ground Truth Ziad Obermeyer, Huong Vu, Alexander Schubert, Saathvik Selvan, and 2 more This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-8802696/v1 This work is licensed under a CC BY 4.0 License Status: Under Review Version 1 posted You are reading this latest preprint version Abstract PCR underpins hundreds of millions of clinical tests each year. It outputs a complex curve that approximates the concentration of a target genetic sequence in a sample. To classify samples as positive, labs then apply simple threshold rules to the curve, discarding most of its rich signal. This approach could introduce misclassifications, but because PCR is considered the gold standard for clinical samples, there is no ground truth against which to measure accuracy. In a set of 24,756 quality-control samples from a federally-certified lab for which ground truth is known, we show that threshold-based approaches to classifications are suboptimal. The lab’s threshold achieved zero false positives but a 4.35% false-negative rate. This was not simply a poor choice of threshold: simulations show no threshold can cut false negatives below 1% without dramatically raising false positives. We train a deep learning model (SPARK) that outperforms threshold rules, and specifically achieves very low false negatives. It performs particularly well on highly diluted samples, the hardest for threshold rules to flag. Finally, in inconclusive clinical samples, SPARK accurately forecasts the results of retesting, using only data from the initial sample. This could reduce retesting by up to 56% while maintaining accuracy. Biological sciences/Computational biology and bioinformatics/Machine learning Health sciences/Health care/Diagnosis/Laboratory techniques and procedures RT-qPCR Machine learning Artificial Intelligence Full Text Additional Declarations There is NO Competing Interest. Supplementary Files SupplementaryInformationSpringerNatureLaTeXTemplate.pdf Supplementary Information Cite Share Download PDF Status: Under Review Version 1 posted 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-8802696","acceptedTermsAndConditions":true,"allowDirectSubmit":false,"archivedVersions":[],"articleType":"Article","associatedPublications":[],"authors":[{"id":593827634,"identity":"37898098-ccdc-44aa-95da-bb07556b666b","order_by":0,"name":"Ziad Obermeyer","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAABCUlEQVRIiWNgGAWjYFACHhiD+RiQSGDgA3PYiNLClgbWwkaCFh4z4rTIt/cefFyZY5PYL33m24OPe9Lk2Nh7DBg+lB3GqYWx51yy4dltaYkz+3K3G854lmPMxnPGgHHGOdxamCVyzCQbtx1O3HCGd5s0z4GKxDaJHANm3jbcWtgkcsx/Nm77n7j/DM8z6T8HKurb5N8YMP/Fo4UHaAtj47YDiRt4eNikGQ7kJLBJ8BgwM+LRIsFzLhnosGTjGWfYzA17DqQZtvGkFRzsOZeOUwsoxD42brOT7e9hfvbgx4FkeX72wxsf/CizxqkFBhwbEGwOgwME1QOBPRKb/QExOkbBKBgFo2DkAACjSFW/7zz53AAAAABJRU5ErkJggg==","orcid":"https://orcid.org/0000-0002-4563-5849","institution":"UC Berkeley","correspondingAuthor":true,"prefix":"","firstName":"Ziad","middleName":"","lastName":"Obermeyer","suffix":""},{"id":593827635,"identity":"53dee26b-0bda-4810-bda1-485b5e047dfe","order_by":1,"name":"Huong Vu","email":"","orcid":"","institution":"University of California, Berkeley","correspondingAuthor":false,"prefix":"","firstName":"Huong","middleName":"","lastName":"Vu","suffix":""},{"id":593827636,"identity":"fcda38ac-dbcc-451e-821a-d1a389167aae","order_by":2,"name":"Alexander Schubert","email":"","orcid":"https://orcid.org/0009-0002-4514-4430","institution":"UC Berkeley","correspondingAuthor":false,"prefix":"","firstName":"Alexander","middleName":"","lastName":"Schubert","suffix":""},{"id":593827637,"identity":"1aad45d2-0238-4c51-8b83-567f26016f4a","order_by":3,"name":"Saathvik Selvan","email":"","orcid":"","institution":"University of California, Berkeley","correspondingAuthor":false,"prefix":"","firstName":"Saathvik","middleName":"","lastName":"Selvan","suffix":""},{"id":593827638,"identity":"82ebeade-99a6-49a3-8573-983115afb1ce","order_by":4,"name":"Petros Giannikopoulos","email":"","orcid":"https://orcid.org/0000-0003-4342-1756","institution":"Innovative Genomics Institute; Universty of California, Berkeley","correspondingAuthor":false,"prefix":"","firstName":"Petros","middleName":"","lastName":"Giannikopoulos","suffix":""},{"id":593827639,"identity":"0cf68aee-1606-4490-8ed3-5fc9d41a1a72","order_by":5,"name":"Ahmed Alaa","email":"","orcid":"https://orcid.org/0000-0001-9936-7141","institution":"University of California, Berkeley; University of California, San Francisco;","correspondingAuthor":false,"prefix":"","firstName":"Ahmed","middleName":"","lastName":"Alaa","suffix":""}],"badges":[],"createdAt":"2026-02-06 04:55:40","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-8802696/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-8802696/v1","draftVersion":[],"editorialEvents":[],"editorialNote":"","failedWorkflow":false,"files":[{"id":104397450,"identity":"171b7acd-5c7c-4c66-8d1d-cb1a1a019381","added_by":"auto","created_at":"2026-03-11 11:48:35","extension":"pdf","order_by":1,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":1260726,"visible":true,"origin":"","legend":"Article File","description":"","filename":"SpringerNatureLaTeXTemplate1.pdf","url":"https://assets-eu.researchsquare.com/files/rs-8802696/v1_covered_f2987cbc-765b-450b-9da8-a7a790158009.pdf"},{"id":103059968,"identity":"1b60d5e0-1e41-45a4-aa04-36bc6a80148b","added_by":"auto","created_at":"2026-02-20 09:47:46","extension":"pdf","order_by":1,"title":"","display":"","copyAsset":false,"role":"supplement","size":527272,"visible":true,"origin":"","legend":"Supplementary Information","description":"","filename":"SupplementaryInformationSpringerNatureLaTeXTemplate.pdf","url":"https://assets-eu.researchsquare.com/files/rs-8802696/v1/6e4177fd7aca0d11e023300a.pdf"}],"financialInterests":"There is \u003cb\u003eNO\u003c/b\u003e Competing Interest.","formattedTitle":"A Deep Learning Approach to Quantitative PCR that Learns from Ground Truth","fulltext":[],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":false,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":true,"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":"nature-portfolio","isNatureJournal":true,"hasQc":false,"allowDirectSubmit":false,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"","title":"Nature Portfolio","twitterHandle":"","acdcEnabled":false,"dfaEnabled":false,"editorialSystem":"ejp","reportingPortfolio":"","inReviewEnabled":true,"inReviewRevisionsEnabled":false},"keywords":"RT-qPCR, Machine learning, Artificial Intelligence","lastPublishedDoi":"10.21203/rs.3.rs-8802696/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-8802696/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"PCR underpins hundreds of millions of clinical tests each year. It outputs a complex curve that approximates the concentration of a target genetic sequence in a sample. To classify samples as positive, labs then apply simple threshold rules to the curve, discarding most of its rich signal. This approach could introduce misclassifications, but because PCR is considered the gold standard for clinical samples, there is no ground truth against which to measure accuracy. In a set of 24,756 quality-control samples from a federally-certified lab for which ground truth is known, we show that threshold-based approaches to classifications are suboptimal. The lab’s threshold achieved zero false positives but a 4.35\\% false-negative rate. This was not simply a poor choice of threshold: simulations show no threshold can cut false negatives below 1\\% without dramatically raising false positives. We train a deep learning model (SPARK) that outperforms threshold rules, and specifically achieves very low false negatives. It performs particularly well on highly diluted samples, the hardest for threshold rules to flag. Finally, in inconclusive clinical samples, SPARK accurately forecasts the results of retesting, using only data from the initial sample. This could reduce retesting by up to 56\\% while maintaining accuracy.","manuscriptTitle":"A Deep Learning Approach to Quantitative PCR that Learns from Ground Truth","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2026-02-20 09:47:41","doi":"10.21203/rs.3.rs-8802696/v1","editorialEvents":[],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"nature-communications","isNatureJournal":true,"hasQc":false,"allowDirectSubmit":false,"externalIdentity":"NCOMMS","sideBox":"Learn more about [Nature Communications](http://www.nature.com/ncomms/)","snPcode":"","submissionUrl":"https://mts-ncomms.nature.com/","title":"Nature Communications","twitterHandle":"","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"ejp","reportingPortfolio":"Nature Communications","inReviewEnabled":true,"inReviewRevisionsEnabled":false}}],"origin":"","ownerIdentity":"519f78f7-1028-49cb-b6c3-4c4d1a492bf4","owner":[],"postedDate":"February 20th, 2026","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"under-review","subjectAreas":[{"id":63191244,"name":"Biological sciences/Computational biology and bioinformatics/Machine learning"},{"id":63191245,"name":"Health sciences/Health care/Diagnosis/Laboratory techniques and procedures"}],"tags":[],"updatedAt":"2026-02-20T09:47:41+00:00","versionOfRecord":[],"versionCreatedAt":"2026-02-20 09:47:41","video":"","vorDoi":"","vorDoiUrl":"","workflowStages":[]},"version":"v1","identity":"rs-8802696","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-8802696","identity":"rs-8802696","version":["v1"]},"buildId":"XKTyCvWXoU3ODBz1xrDgd","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}