Effects of microbiome rare taxa filtering on statistical analysis

Effects of microbiome rare taxa filtering on statistical analysis | 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 Effects of microbiome rare taxa filtering on statistical analysis Quy Xuan Cao, Xinxin Sun, Karun Rajesh, Naga Chalasani, Kayla Gelow, and 4 more This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-34781/v1 This work is licensed under a CC BY 4.0 License Status: Published Journal Publication published 12 Jan, 2021 Read the published version in Frontiers in Microbiology → Version 1 posted You are reading this latest preprint version Abstract Background: Accuracy of microbial community detection in 16S rRNA marker-gene and metagenomic studies suffers from contamination and sequencing errors that lead to either falsely identifying microbial taxa that were not in the sample or misclassifying the taxa of DNA fragment reads. Filtering is defined as removing taxa that are present in a small number of samples and have small counts in the samples where they are observed. This approach reduces extreme sparsity of microbiome data and has been shown to correctly remove contaminant taxa in cultured "mock" datasets, where the true taxa compositions are known. Although filtering is frequently used, careful evaluation of its effect on the data analysis and scientific conclusions remains unreported. Here, we assess the effect of filtering on the alpha and beta diversity estimation, as well as its impact on identifying taxa that discriminate between disease states. Results: The effect of filtering on microbiome data analysis is illustrated on four datasets: two mock quality control datasets where same cultured samples with known microbial composition are processed at different labs and two disease study datasets. Results show that in microbiome quality control datasets, filtering reduces the magnitude of differences in alpha diversity and alleviates technical variability between labs, while preserving between samples similarity (beta diversity). In the disease study datasets, DESeq2 and linear discriminant analysis Effect Size (LEfSe) methods were used to identify taxa that are differentially expressed across groups of samples, and random forest models to rank features with largest contribution towards disease classiffcation. Results reveal that filtering retains significant taxa and preserves the model classification ability measured by the area under the receiver operating characteristic curve (AUC). The comparison between filtering and contaminant removal method shows that they have complementary effects and are advised to be used in conjunction. Conclusions: Filtering reduces the complexity of microbiome data, while preserving their integrity in downstream analysis. This leads to mitigation of the classification methods' sensitivity and reduction of technical variability, allowing researchers to generate more reproducible and comparable results in microbiome data analysis. General Microbiology Filtering Fast Permutation Test Quality Control Microbiome Contaminants Figures Figure 1 Figure 2 Figure 3 Figure 4 Figure 5 Figure 6 Full Text Due to technical limitations, full-text HTML conversion of this manuscript could not be completed. However, the manuscript can be downloaded and accessed as a PDF. Supplementary Files SupplementaryMaterialforFilteringPaper.pdf Cite Share Download PDF Status: Published Journal Publication published 12 Jan, 2021 Read the published version in Frontiers in Microbiology → 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-34781","acceptedTermsAndConditions":true,"allowDirectSubmit":true,"archivedVersions":[],"articleType":"Research","associatedPublications":[],"authors":[{"id":751747,"identity":"dff88939-bb52-4b2c-bb2e-331541938bc2","order_by":0,"name":"Quy Xuan Cao","email":"data:image/png;base64,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","orcid":"https://orcid.org/0000-0002-6204-1305","institution":"University of Montana Missoula College","correspondingAuthor":true,"submittingAuthor":false,"prefix":"","firstName":"Quy","middleName":"Xuan","lastName":"Cao","suffix":""},{"id":751748,"identity":"ca2a3b20-458f-4166-a548-b991333ed10b","order_by":1,"name":"Xinxin Sun","email":"","orcid":"","institution":"Virginia Commonwealth University","correspondingAuthor":false,"submittingAuthor":false,"prefix":"","firstName":"Xinxin","middleName":"","lastName":"Sun","suffix":""},{"id":751749,"identity":"e1cc7a7a-f601-4284-abc0-4a6f8ee358f9","order_by":2,"name":"Karun Rajesh","email":"","orcid":"","institution":"Virginia Commonwealth University","correspondingAuthor":false,"submittingAuthor":false,"prefix":"","firstName":"Karun","middleName":"","lastName":"Rajesh","suffix":""},{"id":751750,"identity":"5c5f4ce1-7738-4bee-8571-c0521b8a3d0b","order_by":3,"name":"Naga Chalasani","email":"","orcid":"","institution":"Indiana University","correspondingAuthor":false,"submittingAuthor":false,"prefix":"","firstName":"Naga","middleName":"","lastName":"Chalasani","suffix":""},{"id":751751,"identity":"4a00cc82-65b6-49b2-a74f-731465b0464c","order_by":4,"name":"Kayla Gelow","email":"","orcid":"","institution":"Indiana University","correspondingAuthor":false,"submittingAuthor":false,"prefix":"","firstName":"Kayla","middleName":"","lastName":"Gelow","suffix":""},{"id":751752,"identity":"0dbfe415-ccb4-41fe-a59c-ab8187f7df46","order_by":5,"name":"Barry Katz","email":"","orcid":"","institution":"Indiana University","correspondingAuthor":false,"submittingAuthor":false,"prefix":"","firstName":"Barry","middleName":"","lastName":"Katz","suffix":""},{"id":751753,"identity":"1f8cbc28-4a17-462c-a922-6a83df418328","order_by":6,"name":"Vijay H. Shah","email":"","orcid":"","institution":"Mayo Clinic Minnesota","correspondingAuthor":false,"submittingAuthor":false,"prefix":"","firstName":"Vijay","middleName":"H.","lastName":"Shah","suffix":""},{"id":751754,"identity":"016197bc-520d-4b51-b1b6-9b4c194b7e35","order_by":7,"name":"Arun J. Sanyal","email":"","orcid":"","institution":"Virginia Commonwealth University","correspondingAuthor":false,"submittingAuthor":false,"prefix":"","firstName":"Arun","middleName":"J.","lastName":"Sanyal","suffix":""},{"id":751755,"identity":"f8e49d30-63d3-4f42-84a0-1cb6b9aa4296","order_by":8,"name":"Ekaterina Smirnova","email":"","orcid":"","institution":"Virginia Commonwealth University","correspondingAuthor":false,"submittingAuthor":false,"prefix":"","firstName":"Ekaterina","middleName":"","lastName":"Smirnova","suffix":""}],"badges":[],"createdAt":"2020-06-12 04:07:16","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-34781/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-34781/v1","draftVersion":[],"editorialEvents":[{"content":"https://doi.org/10.3389/fmicb.2020.607325","type":"published","date":"2021-01-12T19:22:11+00:00"}],"editorialNote":"","failedWorkflow":false,"files":[{"id":1483937,"identity":"8c2ad048-6192-4dbf-be89-27615a1ac94a","added_by":"auto","created_at":"2020-07-02 20:56:32","extension":"png","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":204199,"visible":true,"origin":"","legend":"Heat map and multidimensional scaling plot of MBQC data. Heat map and\nmultidimensional scaling plot of MBQC data. (A) The heat map of 100 observed taxa on the\nlog-scale, with taxa on the x-axis arranged in decreasing abundance order and samples on the\ny-axis arranged by processing institutes. (B) The multidimensional scaling plot of 1016 samples,\ncolored by the processing institutes. Data source: [13].","description":"","filename":"fig1.png","url":"https://assets-eu.researchsquare.com/files/rs-34781/v1/fig1.png"},{"id":1483938,"identity":"d518b997-440b-4f0e-8af4-99b1e3706df4","added_by":"auto","created_at":"2020-07-02 20:56:32","extension":"png","order_by":2,"title":"Figure 2","display":"","copyAsset":false,"role":"figure","size":146084,"visible":true,"origin":"","legend":"Diversity comparison on MBQC and Salter data. (A) Shannon alpha diversity index\nfor the original data and two filtered data, colored by the bioinformatics labs. The horizontal red\ndashed line represents the true Shannon index. (B) Beta diversity multidimensional scaling plots\nof the unfiltered, simultaneous and permutation PERFect filtered data colored by bioinformatics\nprocessing institutes. Data source: [1]. (C) Shannon index for the original data and two filtered\ndata, colored by the dilution levels. (D) Multidimensional scaling plots of the unfiltered and filtered\ndata at different dilution levels, colored by the processing institutes. Data source: [2].\n","description":"","filename":"fig2.png","url":"https://assets-eu.researchsquare.com/files/rs-34781/v1/fig2.png"},{"id":1483939,"identity":"6447a780-3d2b-428a-a8dc-83a495c33f70","added_by":"auto","created_at":"2020-07-02 20:56:32","extension":"png","order_by":3,"title":"Figure 3","display":"","copyAsset":false,"role":"figure","size":53073,"visible":true,"origin":"","legend":"ROC curves of the random forest models from unfiltered and filtered data that are\ndifferentiated by colors. (A) ROC curves from the Alcoholic Hepatitis data [3]. (B) ROC curves\nfrom the IBD data [4].\n","description":"","filename":"fig3.png","url":"https://assets-eu.researchsquare.com/files/rs-34781/v1/fig3.png"},{"id":1483940,"identity":"3697784c-c154-43db-9cd3-6af72df47b0a","added_by":"auto","created_at":"2020-07-02 20:56:33","extension":"png","order_by":4,"title":"Figure 4","display":"","copyAsset":false,"role":"figure","size":347367,"visible":true,"origin":"","legend":"Alcoholic Hepatitis analysis results for Random Forest, LEfSe and DESeq2. (A) Log fold changes for all significant\ntaxa from LEfSe results from unfiltered and filtered data that are differentiated by colors. Taxa that are present in filtered data\nbut are not significant are colored in dark blue. (B) Barchart of log(count+1) for significant taxa from DESeq2 results, colored\nby the disease states. Taxa that are present in filtered data but are not significant are colored in dark blue. (C) Barchart of\nlog(count+1) for common significant taxa between random forest models, LEfSe and DESeq2 results on unfiltered data, colored\nby the disease states. From filtered data, while black taxa are common results with those from unfiltered data, blue taxa are\nnon-significant in DESeq2 results and green taxa are not in the top 60 predictive taxa in the random forest model. Data source:\n[3].\n","description":"","filename":"fig4.png","url":"https://assets-eu.researchsquare.com/files/rs-34781/v1/fig4.png"},{"id":1483941,"identity":"ea8b798b-0a07-4750-8c69-6710052d044e","added_by":"auto","created_at":"2020-07-02 20:56:33","extension":"png","order_by":5,"title":"Figure 5","display":"","copyAsset":false,"role":"figure","size":262386,"visible":true,"origin":"","legend":"IBD analysis results for Random Forest, LEfSe and DESeq2. (A) Log fold changes for all significant taxa from LEfSe\nresults from unfiltered and filtered data that are differentiated by colors. Taxa that are present in filtered data but are not\nsignificant are colored in dark blue. Taxa that are present in unfiltered data but are not significant are colored in dark red.\n(B) Barchart of log(count+1) for significant taxa from DESeq2 results, colored by the disease states. Taxa that are present in\nunfiltered data but are not significant are colored in dark red. Data source: [4].\n","description":"","filename":"fig5.png","url":"https://assets-eu.researchsquare.com/files/rs-34781/v1/fig5.png"},{"id":1483942,"identity":"4124e908-4628-473d-b6ab-e946702bd879","added_by":"auto","created_at":"2020-07-02 20:56:33","extension":"png","order_by":6,"title":"Figure 6","display":"","copyAsset":false,"role":"figure","size":242372,"visible":true,"origin":"","legend":"Comparison of the original data (no filtering), contaminant removal (decontam frequency)\nand filtering (PERFect simultaneous) methods. (A) Heat map of log transformed taxa counts in\ndecreasing abundance order on the x-axis and samples by dilution level on the y-axis for the original\ndata (top panel), data where contaminants are removed using decontam (middle panel) and rare\ntaxa filtered using PERFect (bottom panel). True taxa are colored in green to the left of each\nheatmap; ovals indicate taxa removed by decontam and PERFect methods. (B) Alpha diversity for\nthe three comparisons colored by dilution level and processing institute. (C) Beta diversity PCoA\nBray-Curtis distances plots colored by processing institute and arranged by dilution level (rows)\nand three taxa removal methods (columns). Data source: [2].\n","description":"","filename":"fig6.png","url":"https://assets-eu.researchsquare.com/files/rs-34781/v1/fig6.png"},{"id":13513151,"identity":"2c32a28e-e96c-4b7a-8c27-913a32600eef","added_by":"auto","created_at":"2021-09-16 23:58:30","extension":"pdf","order_by":1,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":880703,"visible":true,"origin":"","legend":"","description":"","filename":"filteringpaperfinal.pdf","url":"https://assets-eu.researchsquare.com/files/rs-34781/v1_covered.pdf"},{"id":1483945,"identity":"40a4b6b6-ecf2-4083-91a3-0db6beaba029","added_by":"auto","created_at":"2020-07-02 20:57:37","extension":"pdf","order_by":1,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":362406,"visible":true,"origin":"","legend":"","description":"","filename":"filteringpaperfinal.pdf","url":"https://assets-eu.researchsquare.com/files/rs-34781/v1_stamped.pdf"},{"id":1483943,"identity":"5020fd0b-b6e8-4933-9bfb-1fb473291ce9","added_by":"auto","created_at":"2020-07-02 20:56:34","extension":"pdf","order_by":1,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":336380,"visible":true,"origin":"","legend":"","description":"","filename":"filteringpaperfinal.pdf","url":"https://assets-eu.researchsquare.com/files/rs-34781/v1/filteringpaperfinal.pdf"},{"id":1483944,"identity":"1740e19e-3baa-4389-a352-0deb9c7d6d85","added_by":"auto","created_at":"2020-07-02 20:56:34","extension":"pdf","order_by":1,"title":"","display":"","copyAsset":false,"role":"supplement","size":90899,"visible":true,"origin":"","legend":"","description":"","filename":"SupplementaryMaterialforFilteringPaper.pdf","url":"https://assets-eu.researchsquare.com/files/rs-34781/v1/SupplementaryMaterialforFilteringPaper.pdf"}],"financialInterests":"","formattedTitle":"Effects of microbiome rare taxa filtering on statistical analysis","fulltext":[{"header":"Full Text","content":" \u003cp\u003eDue to technical limitations, full-text HTML conversion of this manuscript could not be completed. However, the manuscript can be downloaded and accessed as a PDF.\u003c/p\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":"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},"keywords":"Filtering, Fast Permutation Test, Quality Control, Microbiome, Contaminants","lastPublishedDoi":"10.21203/rs.3.rs-34781/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-34781/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003cp\u003eBackground: Accuracy of microbial community detection in 16S rRNA\u0026nbsp;marker-gene and metagenomic studies suffers from contamination and sequencing\u0026nbsp;errors that lead to either falsely identifying microbial taxa that were not in the\u0026nbsp;sample or misclassifying the taxa of DNA fragment reads. Filtering is defined as\u0026nbsp;removing taxa that are present in a small number of samples and have small\u0026nbsp;counts in the samples where they are observed. This approach reduces extreme\u0026nbsp;sparsity of microbiome data and has been shown to correctly remove\u0026nbsp;contaminant taxa in cultured \"mock\" datasets, where the true taxa compositions\u0026nbsp;are known. Although filtering is frequently used, careful evaluation of its effect on\u0026nbsp;the data analysis and scientific conclusions remains unreported. Here, we assess\u0026nbsp;the effect of filtering on the alpha and beta diversity estimation, as well as its\u0026nbsp;impact on identifying taxa that discriminate between disease states.\u0026nbsp;\u003c/p\u003e\u003cp\u003eResults: The effect of filtering on microbiome data analysis is illustrated on four\u0026nbsp;datasets: two mock quality control datasets where same cultured samples with\u0026nbsp;known microbial composition are processed at different labs and two disease\u0026nbsp;study datasets. Results show that in microbiome quality control datasets, filtering\u0026nbsp;reduces the magnitude of differences in alpha diversity and alleviates technical\u0026nbsp;variability between labs, while preserving between samples similarity (beta\u0026nbsp;diversity). In the disease study datasets, DESeq2 and linear discriminant analysis Effect Size (LEfSe) methods were used to identify taxa that are differentially\u0026nbsp;expressed across groups of samples, and random forest models to rank features\u0026nbsp;with largest contribution towards disease classiffcation. Results reveal that filtering retains significant taxa and preserves the model classification ability\u0026nbsp;measured by the area under the receiver operating characteristic curve (AUC).\u0026nbsp;The comparison between filtering and contaminant removal method shows that\u0026nbsp;they have complementary effects and are advised to be used in conjunction. \u003c/p\u003e\u003cp\u003eConclusions:\u0026nbsp;Filtering reduces the complexity of microbiome data, while preserving their\u0026nbsp;integrity in downstream analysis. This leads to mitigation of the classification\u0026nbsp;methods' sensitivity and reduction of technical variability, allowing researchers to\u0026nbsp;generate more reproducible and comparable results in microbiome data analysis.\u003c/p\u003e","manuscriptTitle":"Effects of microbiome rare taxa filtering on statistical analysis","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2020-07-02 20:56:32","doi":"10.21203/rs.3.rs-34781/v1","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}}],"origin":"","ownerIdentity":"54ac6c7c-1ebb-498d-acd9-c292e3108bd9","owner":[],"postedDate":"July 2nd, 2020","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"published-in-journal","subjectAreas":[{"id":140465,"name":"General Microbiology"}],"tags":[],"updatedAt":"2021-07-22T19:22:11+00:00","versionOfRecord":{"articleIdentity":"rs-34781","link":"https://doi.org/10.3389/fmicb.2020.607325","journal":{"identity":"frontiers-in-microbiology","isVorOnly":true,"title":"Frontiers in Microbiology"},"publishedOn":"2021-01-12 19:22:11","publishedOnDateReadable":"January 12th, 2021"},"versionCreatedAt":"2020-07-02 20:56:32","video":"","vorDoi":"10.3389/fmicb.2020.607325","vorDoiUrl":"https://doi.org/10.3389/fmicb.2020.607325","workflowStages":[]},"version":"v1","identity":"rs-34781","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-34781","identity":"rs-34781","version":["v1"]},"buildId":"FbvkV6FR0MCFSLy54lSbu","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}