Evaluation of general 16S ribosomal RNA gene PCR primers ...

The overall coverage and phylum spectrum of 175 primers and 512 primer pairs were evaluated in silico with respect to the SILVA 16S/18S rDNA non-redundant ... SkiptoMainContent Advertisement SearchMenu Menu NavbarSearchFilter ThisissueAllNucleicAcidsResearchSubject:Nucleicacidamplification AllNARJournalsAllJournals MobileMicrositeSearchTerm Search SignIn Issues Sectionbrowse ChemicalBiologyandNucleicAcidChemistry ComputationalBiology DataResourcesandAnalyses GeneRegulation,ChromatinandEpigenetics GenomeIntegrity,RepairandReplication Genomics MolecularBiology NucleicAcidEnzymes RNAandRNA-proteincomplexes StructuralBiology SyntheticBiologyandBioengineering MethodsOnline SurveysandSummaries Database WebServer Advancearticles Submit AuthorGuidelines SubmissionSite OpenAccess Purchase About AboutNucleicAcidsResearch EditorialBoard InformationofReferees AdvertisingandCorporateServices JournalsCareerNetwork Alerts Policies Self-ArchivingPolicy DispatchDates Issues Sectionbrowse ChemicalBiologyandNucleicAcidChemistry ComputationalBiology DataResourcesandAnalyses GeneRegulation,ChromatinandEpigenetics GenomeIntegrity,RepairandReplication Genomics MolecularBiology NucleicAcidEnzymes RNAandRNA-proteincomplexes StructuralBiology SyntheticBiologyandBioengineering MethodsOnline SurveysandSummaries Database WebServer Advancearticles Submit AuthorGuidelines SubmissionSite OpenAccess Purchase About AboutNucleicAcidsResearch EditorialBoard InformationofReferees AdvertisingandCorporateServices JournalsCareerNetwork Alerts Policies Self-ArchivingPolicy DispatchDates Close searchfilter Thisissue AllNucleicAcidsResearch Subject:Nucleicacidamplification AllNARJournals AllJournals searchinput Search AdvancedSearch SearchMenu ArticleNavigation Closemobilesearchnavigation ArticleNavigation Volume41 Issue1 1January2013 ArticleContents Abstract INTRODUCTION MATERIALSANDMETHODS RESULTSANDDISCUSSION CONCLUSIONS AVAILABILITY SUPPLEMENTARYDATA FUNDING ACKNOWLEDGEMENTS REFERENCES ArticleNavigation ArticleNavigation Evaluationofgeneral16SribosomalRNAgenePCRprimersforclassicalandnext-generationsequencing-baseddiversitystudies AnnaKlindworth, AnnaKlindworth 1MaxPlanckInstituteforMarineMicrobiology,MicrobialGenomicsandBioinformaticsResearchGroup,Celsiusstr.1,28359Bremen,2JacobsUniversityBremen,SchoolofEngineeringandSciences,Campusring1,28759Bremen,3RiboconGmbH,D-28359Bremen,Germanyand4DepartmentofMicrobialEcology,UniversityofVienna,Althanstr.14,1090Vienna,Austria Searchforotherworksbythisauthoron: OxfordAcademic PubMed GoogleScholar ElmarPruesse, ElmarPruesse 1MaxPlanckInstituteforMarineMicrobiology,MicrobialGenomicsandBioinformaticsResearchGroup,Celsiusstr.1,28359Bremen,2JacobsUniversityBremen,SchoolofEngineeringandSciences,Campusring1,28759Bremen,3RiboconGmbH,D-28359Bremen,Germanyand4DepartmentofMicrobialEcology,UniversityofVienna,Althanstr.14,1090Vienna,Austria Searchforotherworksbythisauthoron: OxfordAcademic PubMed GoogleScholar TimmySchweer, TimmySchweer 1MaxPlanckInstituteforMarineMicrobiology,MicrobialGenomicsandBioinformaticsResearchGroup,Celsiusstr.1,28359Bremen,2JacobsUniversityBremen,SchoolofEngineeringandSciences,Campusring1,28759Bremen,3RiboconGmbH,D-28359Bremen,Germanyand4DepartmentofMicrobialEcology,UniversityofVienna,Althanstr.14,1090Vienna,Austria Searchforotherworksbythisauthoron: OxfordAcademic PubMed GoogleScholar JörgPeplies, JörgPeplies 1MaxPlanckInstituteforMarineMicrobiology,MicrobialGenomicsandBioinformaticsResearchGroup,Celsiusstr.1,28359Bremen,2JacobsUniversityBremen,SchoolofEngineeringandSciences,Campusring1,28759Bremen,3RiboconGmbH,D-28359Bremen,Germanyand4DepartmentofMicrobialEcology,UniversityofVienna,Althanstr.14,1090Vienna,Austria Searchforotherworksbythisauthoron: OxfordAcademic PubMed GoogleScholar ChristianQuast, ChristianQuast 1MaxPlanckInstituteforMarineMicrobiology,MicrobialGenomicsandBioinformaticsResearchGroup,Celsiusstr.1,28359Bremen,2JacobsUniversityBremen,SchoolofEngineeringandSciences,Campusring1,28759Bremen,3RiboconGmbH,D-28359Bremen,Germanyand4DepartmentofMicrobialEcology,UniversityofVienna,Althanstr.14,1090Vienna,Austria Searchforotherworksbythisauthoron: OxfordAcademic PubMed GoogleScholar MatthiasHorn, MatthiasHorn 1MaxPlanckInstituteforMarineMicrobiology,MicrobialGenomicsandBioinformaticsResearchGroup,Celsiusstr.1,28359Bremen,2JacobsUniversityBremen,SchoolofEngineeringandSciences,Campusring1,28759Bremen,3RiboconGmbH,D-28359Bremen,Germanyand4DepartmentofMicrobialEcology,UniversityofVienna,Althanstr.14,1090Vienna,Austria Searchforotherworksbythisauthoron: OxfordAcademic PubMed GoogleScholar FrankOliverGlöckner FrankOliverGlöckner * 1MaxPlanckInstituteforMarineMicrobiology,MicrobialGenomicsandBioinformaticsResearchGroup,Celsiusstr.1,28359Bremen,2JacobsUniversityBremen,SchoolofEngineeringandSciences,Campusring1,28759Bremen,3RiboconGmbH,D-28359Bremen,Germanyand4DepartmentofMicrobialEcology,UniversityofVienna,Althanstr.14,1090Vienna,Austria *Towhomcorrespondenceshouldbeaddressed.Tel:+494212028970;Fax:+494212028580;Email:[email protected] Searchforotherworksbythisauthoron: OxfordAcademic PubMed GoogleScholar NucleicAcidsResearch,Volume41,Issue1,1January2013,Pagee1,https://doi.org/10.1093/nar/gks808 Published: 28August2012 Articlehistory Received: 12December2011 Accepted: 31July2012 Published: 28August2012 PDF SplitView Views Articlecontents Figures&tables Video Audio SupplementaryData Cite Cite AnnaKlindworth,ElmarPruesse,TimmySchweer,JörgPeplies,ChristianQuast,MatthiasHorn,FrankOliverGlöckner,Evaluationofgeneral16SribosomalRNAgenePCRprimersforclassicalandnext-generationsequencing-baseddiversitystudies,NucleicAcidsResearch,Volume41,Issue1,1January2013,Pagee1,https://doi.org/10.1093/nar/gks808 SelectFormat Selectformat .ris(Mendeley,Papers,Zotero) .enw(EndNote) .bibtex(BibTex) .txt(Medlars,RefWorks) Downloadcitation Close PermissionsIcon Permissions Share Email Twitter Facebook More NavbarSearchFilter ThisissueAllNucleicAcidsResearchSubject:Nucleicacidamplification AllNARJournalsAllJournals MobileMicrositeSearchTerm Search SignIn Close searchfilter Thisissue AllNucleicAcidsResearch Subject:Nucleicacidamplification AllNARJournals AllJournals searchinput Search AdvancedSearch SearchMenu Abstract 16SribosomalRNAgene(rDNA)ampliconanalysisremainsthestandardapproachforthecultivation-independentinvestigationofmicrobialdiversity.Theaccuracyoftheseanalysesdependsstronglyonthechoiceofprimers.Theoverallcoverageandphylumspectrumof175primersand512primerpairswereevaluatedinsilicowithrespecttotheSILVA16S/18SrDNAnon-redundantreferencedataset(SSURef108NR).Basedonthisevaluationaselectionof‘bestavailable’primerpairsforBacteriaandArchaeaforthreeampliconsizeclasses(100–400,400–1000,≥1000bp)isprovided.Themostpromisingbacterialprimerpair(S-D-Bact-0341-b-S-17/S-D-Bact-0785-a-A-21),withanampliconsizeof464bp,wasexperimentallyevaluatedbycomparingthetaxonomicdistributionofthe16SrDNAampliconswith16SrDNAfragmentsfromdirectlysequencedmetagenomes.Theresultsofthisstudymaybeusedasaguidelineforselectingprimerpairswiththebestoverallcoverageandphylumspectrumforspecificapplications,thereforereducingthebiasinPCR-basedmicrobialdiversitystudies. INTRODUCTION Understandingmicrobialdiversityhasbeentheambitionofscientistsfordecades.BecausediversityanalysisbycultivationisproblematicforasignificantfractionofBacteriaandArchaea,culture-independentsurveyshavebeendeveloped.Inthepast,themostcommonlyusedapproachwascloningandsequencingofthe16SribosomalRNAgene(rDNA)usingconservedbroad-rangePCRprimers(1).Withtheadventofmassiveparallelsequencingtechnologies,directsequencingofPCRampliconsbecamefeasible(2–4).In2006,Roche’s454GS20pyrosequencing(5)becamethefirsthigh-throughputsequencingtechnologytobesuccessfullyappliedforlargescalebiodiversityanalysisandwaskeytouncoveringthe‘rarebiosphere’(6).Thecontinuousdevelopmentofthetechnology,offeringreadlengthsofupto1000bpnowadays,furtherimprovedthroughputandresolutionof16SrDNAsequencing(7).Sincethen,additionalhigh-throughputsequencingtechnologieshavebecomecommerciallyavailable.TheattractivenessofIllumina(8)liesinthereducedperbasecostsandcomparativelyhighsequencingdepth(9),despitehavingshortreadlengths.WhilethemajoradvantageofIonTorrent(10)areitsrelativelylowcostandrapidsequencingspeed.Furthermore,PacificBioscience(PacBio)nowemploysthe‘single-moleculereal-time’(SMRT)sequencingtechnology,designedtoachieveaveragereadlengthsofmorethan3000bp(11).ForadetailedreviewofsequencingtechnologiespleaserefertoLomanetal.(12).Thereisnodoubtthattherapiddevelopmentofsequencingtechnologieshasopenedanewdimensioninbiodiversityanalysis,butthediversityoftechnologiesalsoaddscomplexitytotheexperimentaldesignofastudy.ThemostcriticalstepforaccuraterDNAampliconanalysisisstillthechoiceofprimers(4,13).Usingsuboptimalprimers,ormoreprecisely,primerpairs,canleadtounder-representation(14)orselectionagainstsinglespeciesorevenwholegroups(15–17).Usinginappropriateprimersconsequentlyleadstoquestionablebiologicalconclusions(17–19).Inthisstudy,175broadrange16SrDNAprimersand512primerpairswereinvestigatedinsilicowithrespecttooverallcoverageandphylumspectrumforBacteriaandArchaeaaswellasampliconlength.Primersequenceswerecomparedwithall37643716S/18SrDNAsequencesavailableintheSILVAnon-redundantreferencedatabase(SSURefNR)release108(20).Forconsistency,allprimerswererenamedaccordingtotheprimernomenclaturesuggestedbyAlmetal.(21).TwopairsofbacterialPCRprimerswereselectedforempiricalevaluationatthefieldstationHelgolandRoads(NorthSea).Finally,theobtainedresultswerecomparedwithdiversityestimatesfrompreviousmetagenomestudies(22).MATERIALSANDMETHODS Primernomenclature PrimerswererenamedaccordingtoAlmetal.(21).Eachnameiscomposedofsevendash-separatedparts,describing:thetargetgene,therankofthetargetgroup,thetargetgroup,thetargetpositionwithinthegene,theprimerversion,thetargetstrandandthelengthoftheprimer.Forillustration,thesevenpartscomprisingtheprimername‘S-D-Bact-0338-a-A-18’aretobeinterpretedasfollows:Anindicationofthetargetgene.Inthiscase,‘S’forsmallsubunitrDNA(S);AnindicationofthelargesttaxonomicgrouptargetedbythePCRprimer.Forexample,‘D’fordomainlevel;Anabbreviateddescription,limitedtothreetofiveletters,ofthespecifictaxonomicorphylogeneticgrouptargetedbytheprimer.Forexample,‘Bact’forthedomainBacteria;Afour-digitnumberindicatingthe5′positionofthesensestrand.Forexample,‘0338’standsforstartposition338intheEscherichiacolisystemofnomenclature(23);Asinglelowercaseletterindicatingtheversionoftheprobe.Forexample‘a’forafirstversion;AsingleuppercaseletterindicatingwhethertheprobesequenceisidenticaltotheDNAsensestrand(S)ortotheantisense(A)strand;andAnumberindicatingthelengthofthePCRprimer.18basesintheexample.Nomenclatureforinsilicoevaluation Inthisstudy,theterm‘coverage’referstothepercentageofmatchesforagiventaxonomicpath.Taxonomicpathswereconsidered‘notcovered’iftheircoveragewasbelow50%.Theterm‘phylumspectrum’referstothenumberofmatchedphyla.Forexample,ifaprimerorprimerpaircoversthemajorityofallphylaitisdescribedashavinga‘largephylumspectrum’.Selectionofprimers Atotalof175forwardandreverse16SrDNAprimerswerechosenfortheinsilicoevaluation.PrimersequenceswereeitherobtainedfromaliteraturesurveyorprovidedbytheSILVAusercommunityinresponsetoapollontheARB/SILVAmailinglistinJanuary2012(SupplementaryTableS1).Onlyprimerswithanoverallcoverageabove75%foreitherBacteriaorArchaeawereconsideredforprimerpairanalysis.AllprimersareavailableinprobeBase,acomprehensiveonlinedatabaseforrRNA-targetedoligonucleotides,atwww.microbial-ecology.net/probebase/(24).Selectioncriteriaforprimerpairs Primerpairswerechosenaccordingtoannealingtemperatures,overallcoverageofvariableregionsandampliconlength.AnnealingtemperatureswerecalculatedwithOligoCalc(25).Primerpaircombinationswithannealingtemperaturedifferencesoflessthan5°Cwereacceptedaspairs.Suitableprimerpairswereorganizedintothreedifferentgroups(SupplementaryTableS8):GroupShort(GroupS)generates100–400bpfragments.GroupMiddle(GroupM)generates400–1000bpfragments.GroupLong(GroupL)generatesfragments≥1000bp.Atotalof512primercombinationswereevaluated.Thebest30bacterialprimerpairsineachgroupandallarchaealprimerpairswithacombinedoverallcoverage>70%wereanalyzedindetail.Insilicoevaluationofprimersandprimerpairs Primerevaluationwasbasedontwodatasets:Firstly,thenon-redundantSILVAReferencedatabase(releaseSSURef108NR)containing376437sequences.TheSILVASSURef108NRwaspreparedfromallSSUsequenceslongerthan1200bpforBacteriaandEukaryotaandlongerthan900bpforArchaea.SequencesarerequiredtohaveaSINA(26)alignmentqualityvaluebetterthan50(20).RedundantsequenceswereremovedbyclusteringwithUCLUST(27)usinga99%identitycriterion.AsecondSSUdatabasewaspreparedfromtheGlobalOceanSurvey(GOS)(28,29)metagenomesusingtheSILVApipeline.AlignmentwasattemptedwithSINAforallGOSreadsandallsequenceswithanalignmentqualityofatleast30andaminimumlengthof300wereretained,yieldingadatasetof10945sequences.Taxonomicclassificationsforeachreadwereappliedasdescribedbelow.PrimermatchingwasexecutedusingtheprobematchfunctionoftheARBPTserver(30)attwolevelsofstringency,allowingzerooronemismatch,respectively.Foreachprimerandstringencylevelthedatabaseentrieswereseparatedintothreegroups:(i)matches;(ii)mismatches;and(iii)unknown.Thematchstatuswasconsideredtobeunknownifnosequencedatawasavailableatthematchpositionoftherespectiveprimer.Furthermore,onlysequencescorrespondingtotheprimerattheintendedpositionwhereconsideredtobematches.Fromthesenumbers,coveragewascomputedasthematchedfractionofentrieseithermatchesormismatches,excludingentriesforwhichthematchstatuswasunknown.Individualcoverageswerecomputedforalltaxa.Whencomputingthecombinedcoverageofforwardandreverseprimerpairs,anentrywasconsideredtohaveunknownmatchstatusifthematchstatusforeitherofthetwoprimerswasunknown.Likewise,thepairwasonlyconsideredtobeamatchifbothprimersmatchedattheintendedmatchposition.DetailedinformationforeachanalysedprimerandprimerpairareprovidedintheSupplementaryMaterialOnline(singleprimer:SupplementaryTablesS2–S7;primerpairs:SupplementaryTablesS9–S38).AllscriptsandSQLqueriesaswellasdatabasedumpsareavailableonlineatwww.arb-silva.de/download/archive/primer_evaluation.Samplingsiteandcollectionofwatersamples Samplecollectionwascarriedoutaspartofthe‘multiomic’approachoftheMIMAS(MicrobialInteractioninMArineSystems)project(www.mimas-project.de).Surfacewaterwascollectedon11February2009andweeklyfrom31March2009untilOctober2009.Watersamples(totalvolume360l)fromtheKabeltonnesiteatHelgolandRoadsintheNorthSea(54°11.18′N,7°54.00′E)werecollectedatadepthof0.5mandprocessedimmediatelyattheBiologicalStationHelgoland.Thewaterwaspre-filteredthrougha10μmanda3µmpore-sizefilter.Forharvestinga0.2-μm-pore-sizefilterwasused.Ateachtimepoint10land15lofseawaterwerefilteredonto8filtersforgenomicDNAextraction.Allfilterswerestoredat−80°Cuntilfutureusage.DetailscanbefoundinTeelingetal.(22).Inthisstudy,16SrDNApyrotaganalysiswithRoche’s454FLXTitaniumtechnologywasperformedusingsamplesfrom:11February2009,7April2009and14April2009.Resultsfrom16SrDNAdiversityanalysisgainedfrommetagenomestudiesofthesamesamplingdates(22)wereusedforcomparison.DNAextraction GenomicDNAwasdirectlyextractedfromfiltersasdescribedinZhouetal.(31)withthefollowingmodifications:allextractionstepswereperformedwith50µlproteinaseK(10mg/ml),andafterisopropanolprecipitation,pelletednucleicacidswereobtainedbycentrifugationat50000gfor30minatroomtemperature.ThegenomicDNAwasstoredat−20°CuntilPCRamplificationandmetagenomicsequencingwerecarriedout.Amplification Persample,twoseparatePCRreactionswereperformedinordertotesttwobacterialprimerpairsfor16SrDNAamplification.Primerpairswere:(i):S-D-Bact-0341-b-S-17,5′-CCTACGGGNGGCWGCAG-3′(32),andS-D-Bact-0785-a-A-21,5′-GACTACHVGGGTATCTAATCC-3(32);and(ii):S-D-Bact-0008-a-S-16,5′-AGAGTTTGATCMTGGC-3′(33),andS-D-Bact-0907-a-A-20,5′-CCGTCAATTCMTTTGAGTTT-3′(34).Thereactionwascarriedoutin50µlvolumescontaining0.3mg/mlBSA(BovineSerumAlbumin),250µMdTNPs,0.5µMofeachprimer,0.02UPhusionHigh-FidelityDNAPolymerase(FinnzymesOY,Espoo,Finland)and5xPhusionHFBuffercontaining1.5mMMgCl2.ThefollowingPCRconditionswereused:initialdenaturationat95°Cfor5min,followedby25cyclesconsistingofdenaturation(95°Cfor40s),annealing(2min)andextension(72°Cfor1min)andafinalextensionstepat72°Cfor7min.Annealingtemperatureforprimerpair(i)wassetat55°Candfor(ii)at44°C.PCRproductswerepurifiedwithaQiaQuickPCRpurificationkit(QIAGEN,Hilden,Germany).ThequantityandqualityoftheextractedDNAwereanalysedbyspectrophotometryusinganND-1000spectrophotometer(NanoDropTechnologies,Wilmington,DE)andbyagarosegelelectrophoresis.ThePCRproductswerestoredat−20°Cforsequencing.Sequencing ThepyrosequencingreactionswereperformedatLGCGenomicsGmbH,Berlin,Germany.AllsequencingreactionswerebaseduponFLX—Titaniumchemistry(Roche/454LifeSciences,Branford,CT06405,USA;www.454.com)andallmethodswereperformedusingthemanufacturers’protocol.Briefly,genomicDNAfrommetagenomestudies(22)aswellasPCR-amplifiedDNAfragmentswerecheckedforqualityona2%agarosegel.500ngofeachsamplewasthenusedforthesequencinglibrary.Inaminormodificationtotheprotocol,nosizeselectionofthefragmentswasperformed.Thefragmentsweresubjectedtoendrepairandpolishing.AnextraAwasaddedtotheendsofthefragmentsandtheRocheRapidLibraryadaptorswereligatedontothefragmentsasdescribedintheRocheRapidLibraryPreparationManualforGSFLXTitaniumSeries,October2009,Rev.January2010(Roche/454LifeSciences,Branford,CT06405,USA;www.454.com).AftersubsequentemulsionPCRthefragmentlibrarieswereprocessedandsequencedaccordingtotheRocheprotocols.TheresultingsequenceswereprocessedusingthestandardRochesoftwareforbasecalling,trimmingofadaptorsandqualitytrimming(GenomeSequencerFLXSystemSoftwareManualversion2.3,Roche/454LifeSciences,Branford,CT06405,USA;www.454.com).ForPCR-amplifiedDNAfragments,persampletwodistinctPCRreactionsweresequencedon1/8ofpicotiterplate(PTP).RawdatawerestoredasFNAfile.SequencesweresubmittedtoINSDC(EMBL-EBI/ENA,Genbank,DDBJ)withaccessionnumberERP001031.FormetagenomicstwofullPTPspersampleweresequenced.MetagenomesequenceswerepublishedbytheMIMASproject(22)andcanbeobtainedfromINSDCwithaccessionnumberERP001227.Identificationandtaxonomicclassificationof16SrDNAfragments UnassembledsequencereadsfrombothSSUrRNAgenePCRamplicons(pyrotags)andmetagenomesequencingwerepreprocessed(qualitycontrolandalignment)bythebioinformaticspipelineoftheSILVAproject(20).Briefly,readsshorterthan200ntorwithmorethan2%ofambiguitiesormorethan2%ofhomopolymerswereremoved.RemainingreadsfromampliconsandmetagenomeswerealignedagainsttheSSUrDNAseedoftheSILVAdatabaserelease108(www.arb-silva.de/documentation/background/release-108/)(20)usingSINA(26).Unalignedreadswerenotconsideredindownstreamanalysistoeliminatenon16SrDNAsequences.RemainingPCRampliconswereseparatedbasedonthepresenceofalignednucleotidesatE.colipositionsoftherespectiveprimerbindingsitesinsteadofsearchingfortheprimersequencesitself.Thisstrategyisrobustagainstsequencingerrorswithintheprimersignaturesorincompleteprimersignatures.Thisseparationstrategyworksbecausetheampliconsizeofoneprimerpairissignificantlonger,withoverhangsonboth3′and5′site,comparedwiththeampliconofthesecondprimerpair.Withthisapproachtheneedforbarcodingduringcombinedsequencingof16SpyrotagsderivedfromdifferentPCRreactionsonthesamePTPlanewasavoided.FASTAfilesforeachprimerpairoftheseparatedsamplesareavailableonlineatwww.arb-silva.de/download/archive/primer_evaluation.Readsofthefilteredandseparated16Spyrotagdatasetsaswellasmetagenomesweredereplicated,clusteredandclassifiedonasamplebysamplebasis.Dereplication(identificationofidenticalreadsignoringoverhangs)wasdonewithcd-hit-estofthecd-hitpackage3.1.2(www.bioinformatics.org/cd-hit)usinganidentitycriterionof1.00andawordsizeof8.Remainingsequenceswereclusteredagainwithcd-hit-estusinganidentitycriterionof0.98(wordsize8).Thelongestreadofeachclusterwasusedasareferencefortaxonomicclassification,whichwasdoneusingalocalBLASTsearchagainsttheSILVASSURef108NRdataset(www.arb-silva.de/projects/ssu-ref-nr/)usingblast-2.2.22+(http://blast.ncbi.nlm.nih.gov/Blast.cgi)withdefaultsettings.ThefullSILVAtaxonomicpathofthebestBLASThitwasassignedtothereadsifthevaluefor(percentageofsequenceidentity+percentageofalignmentcoverage)/2wasatleast93.Inthefinalstep,thetaxonomicpathofeachclusterreferencereadwasmappedtotheadditionalreadswithinthecorrespondingclusterplusthecorrespondingreplicates(asidentifiedinthepreviousanalysisstep)tofinallyobtain(semi-)quantitativeinformation(numberofindividualreadsrepresentingataxonomicpath).RawoutputdataareavailableintheSupplementaryMaterialinSupplementaryTablesS48–S50.Adjustmentofthetotalnumberofsequencereadstosmallersubsetsbyrandomre-sampling Sequencingdepthmayinfringeonthecomparabilityoftheresultingtaxonomicresolution.Toverifythattheresultsderivedfromthe16Spyrotagswerenotanartefactofdeepsequencing,thetotalnumberof16Spyrotagswasreduceduntilroughlyequalamountsofclassifiedpyrotagsandclassifiedmetagenomereadsremainedforeachsample.Threesubsetsofeach16Spyrotagsamplewereadjustedbywithdrawingequalamountsofsequencesrandomlywithoutreplacement.RawoutputdataareavailableintheSupplementaryMaterialOnline(SupplementaryTablesS51–S52).AnanalogueapproachwasdescribedinGilbertetal.(35).RESULTSANDDISCUSSION Insilicoevaluationof16SrDNAprimers Theoverallcoverageof175singleprimerswasevaluatedforallthreedomainsoflife(SupplementaryTableS1).AdditionallyforBacteriaandArchaeathephylumspectrumwasinvestigatedwithrespecttozeroandonemismatch(SupplementaryTablesS2–S5).Eukaryotaareonlyconsideredondomainlevel(SupplementaryTablesS5–S6).Atotalof122singleprimerspassedthe50%overallcoveragethresholdwith31,51and1primer(s)specificforthedomainArchaea(A),Bacteria(B)andEukaryota(E),respectively.Atone-mismatch-stringencythetotalnumberincreasedto150eligibleprimers.ForArchaea,primersS-D-Arch-0519-a-A-19(A:91.3%,B:0.1%,E:1.0%)andS-D-Arch-0787-a-A-20(A:87.4%,B:7.8%,E:0.0%)standout.ThisisinlinewitharecentstudybyWangandQian(15).ThehighestoverallcoverageandspecificityforthedomainBacteriawasdetectedfortheprimersS-D-Bact-1061-a-A-17(A:2.9%,B:96.4%,E:0.0%)andS-D-Bact-0564-a-S-15(A:16.3%,B:96.0%,E:0.0%).Furthermore,39primersshowrelativelyhighoverallcoverageformorethanonedomain.Forinstance,S-*-Univ-0515-a-S-19(A:54.5%,B:95.4%,E:92.2%)detectsallthreedomainsandS-D-Bact-0787-b-A-20(A:89.9%,B:90.6%,E:0.0%)targetsBacteriaandArchaeaasrecentlyreported(36).Ithaspreviouslybeenasserted(15)thattheprimersS-*-Univ-0789-a-S-18(A:86.1%,B:6.8%,E:0.0%)andS-*-Univ-0906-a-S-17(A:83.7%,B:0.3%,E:76.8%)targetBacteriaandArchaea.Contrarytothis,withonly6.8%and0.3%overallcoverageofthedomainBacteria,but86.1%and83.7%overallcoverageofthedomainArchaea,respectively,ourresultsconfirmtheoriginalintentionofbothprimerstobespecificforthedomainArchaea(37,38).However,ifonemismatchistolerated,S-*-Univ-0789-a-S-18(A:96.0%,B:93.0%,E:0.0%)targetsArchaeaandBacteria.S-*-Univ-0906-a-S-17(A:93.2%,B:49.8%,E:0.0%)stillfailstopassour50%threshold.TheprimersequenceofS-*-Univ-0779-a-S-20(A:0.0%,B:0.0%,E:0.0%)ismisspelledinWangandQian(15).AllowingonemismatchincreasestheoverallcoveragetoA:64.8%,B:6.8%,E:77.6%andindicatesthatthecorrectprimersequencetargetsArchaeaandEukaryota.AdirectcomparisonofourresultswiththestudiesofHuwsetal.(39)andBakeretal.(14)isnotpossible,astheoverallcoverageoftheprimersisnotgiven.Nossaetal.(1)restrictedtheirevaluationtoasinglehabitat.Walteretal.(36)analysedatotalofonlyfourprimers.Inrespecttodetailedphylumcoverage(SupplementaryTablesS2–S5)itshouldbenotedthatthenumbersofsequencespresentinaphylumaffectsthevaluesforphylumcoverage.Ifthemajorityofasmallphylum(e.g.Caldisericawith61sequences)istargeted,thecoveragewillbehigherthanforamemberrichphylum(e.g.Firmicuteswith84910sequences).Similareffectsoccurforphylainwhichonlyasmallnumberofsequencescontainsequenceinformationattheprimerpositionofinterest.Insilicoevaluationofprimerpairs Whencombiningforwardandreverseprimers,thebiasofsingleprimerscanaccumulate.Tominimizetheoverallbias,primerswithsimilaroverallcoverageandphylumspectrummustbeused.Usingthe75%overallcoveragecriterion,86singleprimersqualifyforprimerpairanalysis.Inordertogetsuitablecombinationsforthedifferentsequencingtechnologies,primerpairswereorganizedintothreegroupsaccordingtotheirampliconlength(SupplementaryTableS8).GroupS(mall)couldbeofparticularinterestforIllumina(8)andIonTorrent(10)sequencing.PrimerpairsofGroupM(iddle)aresuitableforRoche’s454(40)technology.GroupL(arge)primerpairsareusefulforsequencingmethodssuchasPacBio(11)aswellasforcreatingclassicalclonelibraries.Atotalof512primercombinationswereevaluated.Again,thefocusofthisevaluationwasArchaeaandBacteria.Eukaryotaareonlyconsideredondomainlevel.AssumingthatastandardPCRcantolerateuptotwomismatchesbetweentheprimeranditstarget(1),resultswithonemismatcharealsotakenintoaccount.However,itshouldbenotedthataprimermismatchcanresultinabiasedpictureofthebacterialdiversity(41)andpreferentialamplificationmightleadtounder-representationofimportantmembersofacommunity(14,41).InsilicoevaluationofprimerpairssuitableforIlluminaandIonTorrentsequencing(GroupS) Only12archaealprimerpairshaveanoverallcoverageabove70%.Thebestresultswithanoverallcoverageof76.8%areobtainedwithS-D-Arch-0349-a-S-17/S-D-Arch-0519-a-A-16(A:76.8%,B:0.0%,E:0.0%)(SupplementaryTableS9).Thispairgeneratesanampliconlengthof185bpwhichspansthehypervariable(HV)regionthree.Theevaluationrevealedthatitmissesfiveoutofeightphyla:AncientArchaealGroup(AAG),GoC-Arc-109-D0-C1-M0,Korarchaeota,MarineHydrothermalVentGroup2(MHVG-2)andNanoarchaeota.Theremainingthreearchaealphylaaredetected(Crenarchaeota,MarineHydrothermalVentGroup1(MHVG-1)andEuryarchaeota).Withonemismatchallowed,overallcoverageforArchaeaincreasestoA:91.0%,B:0.0%,E:0.1%nowcoveringadditionallyKorarchaeotaandMHVG-2(SupplementaryTableS10).However,inthecaseofKorarchaeotadetailedanalysisoftheprimertargetpositionrevealeda3′endmismatchoftheforwardprimer,whichisknowntoaffectamplification.NanoarchaeotaandAAGshowthreemismatches.Moreover,PCRhastotolerateuptofourmismatchesoftheforwardprimertoamplifyGoC-Arc-109-D0-C1-M0.Insummary,S-D-Arch-0349-a-S-17/S-D-Arch-0519-a-A-16generatesshortamplicons,hasacomparativelyhighoverallcoveragebydetectinguptofouroutofeightarchaealphylaandexcellentdomainspecificity.Hence,thisprimerpairshowsthemostpromisingresultsforIlluminaandIonTorrentsequencing.ForBacteria,theprimerpairS-D-Bact-0341-b-S-17/S-D-Bact-0515-a-A-19(A:0.0%,B:91.2%,E:0.0%)hasthehighestoverallcoverage(SupplementaryTableS11).Detailedanalysisrevealsthat10phylaarenotdetected(Armatimonadetes,Chlamydiae,Caldiserica,Hyd24-12,GOUTA4,Kazan-3B-28,SM2F11,aswellasCandidatedivisionsWS6,OP11,TM7andOD1).IfonemismatchistoleratedsomeArchaea(A:44.6%,B:96.7%,E:0.2%)aswellassevenadditionalphylaaredetected(SupplementaryTableS12),butamplificationofCandidatedivisionsOP11andWS6aswellasArmatimonadetesremainsunlikely.Inallthreecases,themismatchpositionoftheforwardprimerislocatedatthe3′end.ForCandidatedivisionsOP11andWS6,thereverseprimerwouldneedtotoleratethreemismatches.ThesefindingsareinlinewiththeconclusionsofBakerandCowan(42),whoclaimthatnodomain-specificprimerexistsorcanbedesignedthatmatchesallbacterial16SrDNAsequences.ThebestcandidateforthedomainBacteriaisS-D-Bact-0564-a-S-15/S-D-Bact-0785-b-A-18.ThisprimerpairhasaslightlyloweroverallcoverageforBacteria(A:14.6%,B:89.0%,E:0.0%)comparedwiththepreviouscandidatebutonlyfailstodetectfourbacterialphyla(Chloroflexi,Elusimicrobia,BHI80-139andCandidatedivisionOP11).Withoneallowedmismatch(A:57.1%,B:95.2%,E:0.0%),onlyCandidatedivisionOP11sequencesremainundetectedduetoa3′endmismatchofbothprimers.Pleasenotethatonemismatchmayalsoleadtoamplificationofarchaeal16SrDNAsequences.BasedonthepromisingphylumspectrumweareinfavourofthisprimerpairincomparisontothepreviousdescribedS-D-Bact-0341-b-S-17/S-D-Bact-0515-a-A-19.Insummary,S-D-Bact-0564-a-S-15/S-D-Bact-0785-b-A-18generatesanampliconof253bpcoveringthefourthHVregionandsatisfieswithahighoverallcoverageandreasonablygooddomainspecificity.Hence,itisrecommendedforBacteria.TwoprimerpairstargetthedomainsBacteriaandArchaea:S-D-Arch-0519-a-S-15/S-D-Bact-0785-b-A-18(A:88%,B:89.1%,E:0.7%)andS-D-Arch-0519-a-S-15/S-D-Bact-0785-a-A-21(A:86.5%,B:87.1%,E:0.0%).Withinthebacterialdomain,thosetwoprimerpairscover49outof59phyla.ThecoverageforChlamydiae,Caldiserica,Chloroflexi,SM2F11,Kazan-3B-28,BHI80-139andCandidatedivisionsWS6,OP11,TM7andOD1isbelow50%.Ifonemismatchistolerated,sevenadditionalphylaaredetectedandoverallcoverageincreasesforS-D-Arch-0519-a-S-15/S-D-Bact-0785-b-A-18(A:94.9%,B:95.1%,E:1.6%)andS-D-Arch-0519-a-S-15/S-D-Bact-0785-a-A-21(A:94.6%,B:94.8%,E:0.7%).AmplificationofCandidatedivisionsWS6,TM7andOP11remainsunlikely.ThemismatchpositionofS-D-Arch-0519-a-S-15islocatedatthe3′endincaseofCandidatedivisionsWS6andTM7.ForCandidatedivisionOP11,bothreverseprimersshowa3′endmismatch.ForArchaea,eachprimerpairfailstodetectfouroutofeightphyla(AAA,MHVG-1andMHVG-2andNanoarchaeota),whichisreducedtoone(Nanoarchaeota)ifonemismatchisallowed.ThecontinuousfailureofprimerstodetectNanoachaeotaisnotsurprising,duetothemajorityofArchaea-specificprimersbeingdesignedpriortothediscoveryoftheNanoarchaeota(14).DetailedanalysisofthemismatchpositionsrevealsoneinternalmismatchforAAA,MHVG-1andMHVG-2butthreemismatchesforNanoarchaeota.AdditionofNanoarchaeota-specificprimers(43)isrecommended.PreviousevaluationshowedS-P-Nano-0008-a-S-16andS-P-Nano-1390-a-A-17tobehighlyspecificforNanoarchaeota(SupplementaryTableS2).Notethattheseprimersgeneratealmostfull-lengthsequences.Insummary,bothprimerpairscanberecommendedforamplification.TheygenerateampliconsspecificforBacteriaandArchaeawithanaveragelengthof278bpthatspanstheHVregionfour.InsilicoevaluationofprimerpairssuitableforsequencingtechnologieslikeRoche’s454(GroupM) Noarchaeal-specificprimerpairachievesafullphylumspectrum(SupplementaryTableS15).S-D-Arch-0519-a-S-15/S-D-Arch-1041-a-A-18(A:76.6%,B:0.0%,E:0.0%)showsthebestresultswithrespecttoarelativelyhighoverallcoveragecoupledwithahighdomainspecificity.Thisprimerpaircoverstwooutofeightphyla(CrenarchaeotaandEuryarchaeota),butthephylumspectrumincreasesremarkablytosixdetectedphylaifonemismatchisallowed(A:92.8%,B:0.0%,E:0.0%).Detectionofthefouradditionalphyla(AAG,Korarchaeota,MHVGIandMHVGII)islikelyduetoamiddlemismatchpositioninthereverseprimer.AmplificationofGoC-Arc-109-D0-C1-M0andNanoarchaeaotaremainschallengingduetomorethanonemismatch.Insummary,S-D-Arch-0519-a-S-15/S-D-Arch-1041-a-A-18isthemostsuitableprimerpairwitha540bpampliconspanningHVregions4-6andexcellentdomainspecificity.ThefrequentuseofHVregionsixindiversityanalysismakesthispairparticularlyinterestingforcomparativeanalysis(35,44,45).ForthedomainBacteria,severaldomain-specificprimerpairsattainhighoverallcoverage,but27outof30failtodetectmorethan10phyla(SupplementaryTableS17).ThethreebestpairsareS-D-Bact-0341-b-S-17/S-D-Bact-1061-a-A-17(A:0.0%,B:91.9%,E:0.0%),S-D-Bact-0564-a-S-15/S-*Univ-1100-a-A-15(A:8.0%,B:92.7%,E:0.0%)andS-D-Bact-0341-b-S-17/S-D-Bact-0785-a-A-21(A:0.5%,B:86.2%,E:0.0%).Althoughthefirsttwoshowhigheroverallcoverage,thelatterexhibitsalargerphylumspectrum.S-D-Bact-0341-b-S-17/S-D-Bact-0785-a-A-21onlyfailstodetectsevenbacterialphyla(Hyd24-12,GOUTA4,Armatimonadetes,Chloroflexi,BHI80-139andCandidatedivisionsOP11andWS6).Ifonemismatchistolerated(A:64.6%,B:94.5%,E:0.1%),amplificationoffouradditionalphylaislikely(Chloroflexi,BHI80-139,Hyd24-12andGOUTA4).However,somearchaealsequencesarealsodetected.DetailedanalysisrevealsthatonlythecoverageforCandidatedivisionOP11remainsbelowthe50%threshold(SupplementaryTableS18).Besidesfourmismatchesforthereverseprimer,themismatchpositionsinbothprimersarelocatedtowardsthe3′end.Moreover,amplificationofArmatimonadetesandCandidatedivisionWS6isunlikelyduetothe3′endmismatchpositionoftheforwardprimer.Althoughnotcoveringthecompletephylumspectrum,thepairS-D-Bact-0341-b-S-17/S-D-Bact-0785-a-A-21showsthebestcombinationofdomainandphylumcoverageandcanthusberecommendedfor464bpampliconscoveringtheHVregions3–4.S-D-Bact-0785-a-S-18/S-*-Univ-1392-a-A-15(A:72.3%,B:74.1%,E:0.0%)qualifiesasasuitableprimerpairforBacteriaandArchaea.WithnomismatchesitonlyfailstodetectNanoachaeotaandexpandstofullarchaealphylumspectrumifonemismatchistolerated.Detailedanalysisrevealedthatnoneofthemismatchpositionsarelocatedtowardsthe3′end,whichshouldallowamplification.ForBacteria,anoverallcoverageof76.3%isachievedbutthispairfailstodetectninephyla(Chloroflexi,SM2F11,HDB-SIOH1705,BD1-5,EM19,BHI80-139,CandidatedivisionsOP11,SR1,OD1aswellasEpsilonproteobacteria).Allowingonemismatchresultsinanincreasedoverallcoverage(A:79.0%,B:86.1%,E:1.3%)andtheadditionaldetectionofsixphyladuetointernalmismatches.OnlythecoverageofHDB-SIOH17005,SM2F11andCandidatedivisionOP11remainsbelowthe50%threshold.Insummary,withanampliconlengthof608bpanddetectionofHVregions5–8thisprimerpairqualifiestotargetBacteriaandArchaea.Thisdetailedevaluationalsodemonstratesthatreverseandforwardprimerswithindividualhighcoveragedonotautomaticallyqualifyasanoptimalprimerpair.Forinstance,S-D-Bact-0347-a-S-19(A:0.0%,B:86.1%,E:0.0%)andS-D-Bact-0785-a-A-19(A:8.5%,B:86.4%,E:0.0%)havebeendesignedandapprovedbytheHumanMicrobiomeProjectforanalysingtheforegutmicrobiome(1).Basedonpromisingresultswithinthehumanhabitat,theysuggestedthatthisprimerpairmaybeagoodcandidatetoaccessthebacterialdiversityinanyhabitat(1).However,ourevaluationrevealsaloweroverallcoverageofA:0.0%,B:76.5%,E:0.0%anddetectionofonly25outof59bacterialphylaiftheyactasaprimerpair.Evenifonemismatchisallowed(A:0.0%,B:90.6%,E:0.0%)thisprimerpairstillfailstodetect17phyla(Armatimonadetes,Chlamydiae,Dictyoglomi,Planctomycetes,Verrucomicrobia,Spirochaetes,Lentisphaerae,HDB-SIOH1705,LD1-PA38,NPL-UPA2,Hyd24-12andSM2F11,aswellasCandidatedivisionsOP11,WS6,BRC1,OD1,WS3andOP3).InsilicoevaluationofprimerpairssuitableforsequencingtechnologiessuchasPacBioSMRTorclassicalclonelibraries(GroupL) Forfragments>1000baseswecouldnotfindanarchaealprimerpairwithbothanoverallcoverageofover70%andasatisfyingphylumspectrum(SupplementaryTableS21).Themajoritydetectsonlythetwosequence-richphyla,CrenarchaeotaandEuryarchaeota.S-D-Arch-0349-a-S-17/S-*-Univ-1392-a-A-15(A:65.8%,B:0.0%,E:0.0%)hasthehighestoverallcoverage.Detailedanalysisrevealedthatthispairfailstodetectsixoutofeightphyla(AAG,GoC-Arc-109-D0-C1-M0,Korarchaeota,MHVG-1,MHVG-2andNanoarchaeota)(SupplementaryTableS21).Althoughperformanceincreasesslightlywhenonemismatchisallowed(A:76.0%,B:0.0%,E:0.1%),thecoverageforthreephyla(AAG,GoC-Arc-109-D0-C1-M0andNanoarchaeota)remainsbelow50%(SupplementaryTableS22).Inaddition,a3′mismatchoftheforwardprimerhampersamplificationofKorarchaeota.Insummary,thisprimerpaircannotberecommended.SimilarresultsareobtainedfortheotherarchaealprimerpairsofGroupL.Thebacterialprimerpairsshowmoresatisfyingresults(SupplementaryTableS23).S-D-Bact-0008-c-S-20/S-D-Bact-1391-a-A-17(A:0.1%,B:78.0%,E:0.0%)hasahighoverallcoverageanddetects55outof59phyla.Thefourphylawithbelow-thresholdcoverageareChlamydiae,WCHB1-60,CandidatedivisionSR1andOP11.Ifonemismatchisallowed,overallcoverageincreasestoA:0.1%,B:86.2%,E:0.0%andCandidatedivisionOP11isnowlikelytobedetectedduetoaninternalmismatch.S-D-Bact-0008-c-S-20/S-D-Bact-1046-a-A-19(A:0.0%,B:81.3%,E:0.0%)achievesthehighestoverallcoveragebutfailstodetecteightphyla(S2R-29,SM2F11,Chlamydiae,Thermotogae,WCHB1-60,Kazan-3B-28,EM19,CandidatedivisionOP11andEpsilonproteobacteria).Remarkably,thisismostlycompensatedifonemismatchisallowed.However,amplificationofsomesequencesbelongingtoCandidatedivisionOP11andWHCBI-60isunlikelydueto3′endmismatches.Moreover,thereverseprimerfailstodetectSM2F11duetotwomismatchesofwhichoneislocatedtowardsthe3′end.Chlamydiaeremainsundetectedduetothreeinternalmismatchesoftheforwardprimer.ThepromisingresultsandexcellentdomainspecificityofbothprimerpairsaredepreciatedbythefactthattheyonlyspanHVregions1–6and1–8,respectively.Nevertheless,ifanampliconlengthof<1400bpissufficientweareinfavourofbothprimerpairs.Fornearlyfull-lengthsequences(>1400bp)werecommendS-D-Bact-0008-a-S-16/S-D-Bact-1492-a-A-16(A:0.2%,B:77.1%,E:0.0%).Thisdomain-specificprimerpairspansHVregions1–9andcovers52outof59bacteriaphyla.Themissingphylaare:GAL08,Kazan-3B-28,Chlamydiae,Dictyoglomi,WCHB1-60,MVP-21andCaldiserica.Onemismatch(A:0.2%,B:86.8%,E:0.0%)allowsadditionaldetectionofCaldisericaandDictyoglomiduetoaninternalmismatch.Theremainingfivephylahaveeithermorethantwomismatchesor,incaseofChlamydiae,theforwardprimerhasa3′endmismatch.Inthepast,S-D-Bact-0008-a-S-16/S-D-Bact-1492-a-A-16,whichiscommonlyknownasGM3/GM4,hasbeenintensivelyusedforclonelibrary-basedstudiesfromdifferenthabitats(46–48).Thusplentyofdataforcomparativeanalysisareavailable.However,thehighnumberofsequencesoriginallyobtainedwiththeGM3/GM4pairisalsolikelytohaveartificiallyinflatedthecoveragevaluesweobtained.Ideally,sequencesobtainedwithagivenprimershouldbeexcludedwhenevaluatingthatsameprimer.Insilicore-evaluationofprimerpairsusingaPCRfreemetagenomedatabase Themajorityofthesequencesinspecialized16S/18SrDNAdatabasessuchasSILVA(20),greengenes(49)orRDPII(50)arearesultofpriorPCRamplification.Inordertocalibrateourpreviousanalysis,re-evaluationoftheresultsusingthepubliclyavailableGlobalOceanSampling(GOS)databasewasperformed.TheinitialGOSdatasetconsistedof6.3billionbpofSangersequencereads(28)andhasrecentlybeenaugmentedbysamplesfromtheAtlanticandIndianOceans(51).Althoughitislimitedtothemarinehabitat,itisthemostcomprehensivedatasetthatprovidesareasonableamountofrelativelylongfragmentsnecessaryforprimerevaluation.Atotalof1068516S/18SrDNAsequenceswereextractedfromtheGOSdataset.95%ofthereadsrangebetween900bpand1200bpinlength;theaveragelengthwas1053bp.However,thebacterialfractionwasdominant,consistingof9965sequences,comparedwithonly290archaealand439eukaryotic16Sand18Ssequences,respectively.ThustheresultsforArchaeaandEukaryotaareuncertainandshouldonlybeseenascomplementaryinformation.Inadditiontothelimitednumberofsequences,onlyasubsetofphylaispresent.Forexample,forArchaea288sequencesbelongtoCrenarchaeota(63sequences)andEuryarchaeota(225sequences).TheremainingtwosequencescouldbeassignedtoAAGandMHVG-1,respectively.ForKorarchaeota,GoC-Arc-109-D0-C1-M0,MHVG-2andNanoarchaeota,nosequencesarepresent.ForthedomainBacteria,the9956readsspan28outof59phyla.ThemajoritybelongtoActinobacteria(1006sequences),Bacteroidetes(785sequences),Cyanobacteria(805sequences)andProteobacteria(6655sequences).OthermemberrichphylasuchasFirmicutes(167sequences)andAcidobacteria(29sequences)areonlypresentinlownumbers.Thelackofafullphylumspectrumclearlylimitsthere-evaluationandpreventsdirectcomparisonswithourpreviousresults.Themuchlowerandalsovaryingnumberofsequencesintherespectivetargetregionsaffectstheresultsaswell.FurthermoreprimerpairsofGroupLhadtobeexcludedfromthere-evaluationduetothelackofsufficientnumbersoflongsequences.InthepreviousevaluationforGroupS,thearchaealprimerpairS-D-Arch-0349-a-S-17/S-D-Arch-0519-a-A-16(A:76.8%,B0.0%,E:0.0%)wasproposedasasuitablepairforampliconsequencingof<400bases.Re-evaluationbasedontheGOSdatasetagainyieldedthehighestoverallcoverage(A:74.5%,B:0.0%,E:1.2%)andexcellentdomainspecificity.TherecommendedbacterialprimerpairS-D-Bact-0564-a-S-15/S-D-Bact-0785-b-A-18(A:0.0%,B:83.4%,E:0.0%)alsoperformswell.Toleratingonemismatchstillconfirmsdomainspecificity(A:10.6%,B:86.2%,E:0.0%).Unfortunately,detailedcomparisononphylumlevelproveddifficult.Forexample,withintheSILVAdatabase,84910Firmicutessequencesofsufficientlengtharepresentand91.8%ofthesearecoveredbyS-D-Bact-0564-a-S-15/S-D-Bact-0785-b-A-18.UsingtheGOSdataset,onlytwosequencesfromFirmicutesareavailable.Promisingtrendscouldalsobeobservedforthetwoprimerpairstargetingboth,ArchaeaandBacteria.Inparticular,S-D-Arch-0519-a-S-15/S-D-Bact-0785-b-A-18standsoutwithhighoverallcoverage(A:76.5%,B:83.4%,E:1.9%),whichincreasesslightlyifonemismatchisallowed(A:81.8%,B:86.5%,E:1.9%).ForGroupM,only32sequencesofsufficientlengthwereavailabletore-evaluatetherecommendedarchaealprimerpairS-D-Arch-0519-a-S-15/S-D-Arch-1041-a-A-18.ThustheArchaeaprimerpairswereexcludedfromfurthervalidation.Withonaverage2600availablebacterialsequencesforre-evaluatingGroupM,theconditionswereslightlybetter.Asinthepreviousevaluation,severalprimerpairsshowhighoverallcoverage:S-D-Bact-0564-a-S-15/S-*-Univ-1100-a-A-15provesitssuitabilitywithahighdomain-specificandoverallcoverage(A:0.0%,B:76.2%,E:0.0%).OverallcoverageforBacteriaincreasesupto80.2%,ifonemismatchistolerated(A:2.3%,B:80.2%,E:0.0%).Incontrast,S-D-Bact-0341-b-S-17/S-D-Bact-1061-a-A-17(A:0.0%,B:58.9%,E:0.0%)failstomatchthepreviousresults,whichcouldbeaconsequenceofthespecificdataset.Evenallowingonemismatchdoesnotachievesatisfyingresults(A:0.0%,B:64.8%,E:0.0%).Atfirstglance,similarresultswereobtainedforS-D-Bact-0341-b-S-17/S-D-Bact-0785-a-A-21(A:0.0%,B:43.1%,E:0.0%).However,consideringonemismatchtheoverallcoveragesignificantlyincreasedtoA:58.2%,B:70.9%,E:0.0%.There-evaluationoftheprimerpairsbasedontheGOSdataset(SupplementaryTablesS27–S38)showsthat,despitetherelativelylargedatasetsize,itstilllacksresolutionpower,especiallywhenconsideringaspecificgene.Unfortunately,thedataobtainedbyotherlargescaleprojects,suchastheEarthMicrobiomeProject(52),isoflittleuseforprimerevaluationduetotheircosteffective,butlength-limitedsequencingstrategy.Duetotheinherentriskofcreatingchimericsequenceswewouldnotconsiderassemblyasolutiontothislimitation.ShouldtheerrorrateoflongreadsequencingtechnologiessuchasPacBiobesignificantlyreduced,datafrommetagenomicstudiesrelyingonthesetechnologieswouldbecomeavaluableresourceforrevisitingtheprimersensitivityissue.Insummary,ifasufficientamountformetagenomic16SrDNAsequenceswereavailable,thepreviousprimerpairrecommendationscouldbeconfirmed.ExperimentalevaluationoftheprimerpairS-D-Bact-0341-b-S-17/S-D-Bact-0785-a-A-21 TheprimerpairS-D-Bact-0341-b-S-17/S-D-Bact-0785-a-A-21(GroupM)wasappliedtoDNAextractedfromatimeseriesofthreemarineenvironmentalsamplesatHelgolandRoads.Forsimplification,wewillrefertotheobtainedreadsas‘16Spyrotags’.InthecourseoftheMIMASproject,metagenomicanalysiswasperformedusingmarinesamplesfromthesamesiteandtimepoints(22).Theresultsfromthemetagenomic-baseddiversitystudiesareusedtoevaluatetheaccuracyofeachprimerpairbycomparingthetaxonomicclassifications.Onaverage,59700sequenceswereobtainedpersamplingoccasion,ofwhich52400couldbeassignedas16Spyrotags(88.4%)(SupplementaryTableS39).Therelativelyhighlossisduetothestringentqualitychecksusedfortheidentificationandtaxonomicclassificationof16SrDNAfragments.Incontrast,metagenomeanalysisresultedonaveragein2109000sequences(22)persamplingoccasion,butonly1600sequences(0.1%)qualifiedas16SrDNAgenefragments.Theresultsofthe16SpyrotaganalysisshowthatthebacterialcommunityisdominatedbyAlphaproteobacteria,BacteroidetesandGammaproteobacteria(Figure1AandSupplementaryTableS40).Accordingtotheinsilicoevaluation,forprimerpairS-D-Bact-0341-b-S-17/S-D-Bact-0785-a-A-21highcoverageofthesethreegroupsareexpected(Bacteroidetes:89.2%,Alphaproteobacteria:81.4%,Gammaproteobacteria:90.6%).Allowingonemismatchtheoverallcoverageincreasestoupto95%foreachgroup.Theresultsfromthe16Spyrotagsalsorevealedasuccessionoftherelativeabundances.Bacteroidetespeakedon7April2009,butwerestillabundanton14April2009.ForAlphaproteobacteriamoresequencescouldbedetectedinwinteron11February2009.Incontrast,therelativeabundanceofGammaproteobacteriaincreasedon14April2009.Thesametrendswereobservedinthemetagenomes(Figure1BandSupplementaryTableS41)(22).Toverifythattheresultsderivedfromthe16Spyrotagsarenotanartefactofdeepsequencing,thetotalnumberofreadswasadjustedtosmallersubsetsofaround2000sequencesbyrandomre-sampling.Detailedanalysisofthesere-sampledsubsetsconfirmedtheresults(SupplementaryTableS42). Figure1.OpeninnewtabDownloadslideTaxonomicdistributionof16SrRNAgenesequencesgainedfromatimeseriesofthreedifferentsurfacewatersamplesatHelgolandRoadsintheNorthSea,(A)16SpyrotagsgeneratedfromPCRandsequencedwithRoche’s454pyrosequencing(relativeabundance,percentageoftotalcounts)(B)16Ssequencesgainedfrommetagenomestudies(relativeabundance,percentageoftotalcounts).Figure1.OpeninnewtabDownloadslideTaxonomicdistributionof16SrRNAgenesequencesgainedfromatimeseriesofthreedifferentsurfacewatersamplesatHelgolandRoadsintheNorthSea,(A)16SpyrotagsgeneratedfromPCRandsequencedwithRoche’s454pyrosequencing(relativeabundance,percentageoftotalcounts)(B)16Ssequencesgainedfrommetagenomestudies(relativeabundance,percentageoftotalcounts).16Spyrotaganalysisprovidesanenhancedresolutionuptothegrouporgenuslevel.Sixrelativelyabundanttaxonomicgroupsandgenera(Formosa,Polaribacter,SAR11cladesurface1,NAC11-7lineage,ReinekeaandSAR92clade)havebeenexaminedindetail(SupplementaryFigureS1AandSupplementaryTableS43).NoteworthyistheFormosapeakon7April2009andthepresenceofReinekeaonlyon14April2009.Bothresultsweresupportedbydiversitystudiesfromthecorrespondingmetagenomes(SupplementaryFigureS1BandSupplementaryTableS44).Again,there-sampled16Spyrotagsubsetsconfirmedthattheresultsarenotanartefactofdeepsequencing(SupplementaryTableS45).Inaddition,itisinterestingtonotethatcorrespondingmetaproteomestudiesdescribedinTeelingetal.(22)reflectthesamesuccessionofthebacterialcommunityontheproteinlevel.Consideringtheinsilicoevaluation,S-D-Bact-0341-b-S-17/S-D-Bact-0785-a-A-21shouldfailtodetectSAR11cladesurface1(0.7%).However,experimentalevaluationclearlyshowsthattheprimerpairisabletoamplifythistaxonomicgroup.Thiscanbeexplainedbytheincreasedcoverageofupto97%ifonemismatchisallowed.Acloserlookattheprimertargetpositionofthereverseprimerrevealsaninternalmismatchpositiontowardsthe5′end.TheresultsdemonstratethatS-D-Bact-0341-b-S-17/S-D-Bact-0785-a-A-21providesagoodrepresentationofthebacterialdiversitydowntogenusandgrouplevelandillustratesthataninternalmismatchtowardsthe5′endcanbetoleratedbystandardPCR.Totesttheassumptionthatasuboptimalprimerpairmightresultinabiasedpictureofthebacterialdiversity,S-D-Bact-0008-a-S-16/S-D-Bact-0907-a-A-20wasappliedtothesamesamples.Thisprimerpairwaschosenduetoitsrelativelyhighoverallcoverage(A:0.0%,B:75.1%,E:0.0%)butdistinctlylowerphylumspectrum.Basedontheinsilicoevaluationitshouldfailtodetect18bacterialphyla(Aquificae,BD1-5,BHI80-139,Chlamydiae,Dictyoglomi,EM19,Lentisphaerae,SM2F11,Thermotogae,Tenericutes,Verrucomicrobia,WCHB1-60andCandidatedivisionsTM7,WS6,OD1,SR1andOP11).WithrelativelyhighcoverageofBacteroidetes(77.6%),Alphaproteobacteria(71.3%)andGammaproteobacteria(80.5%)insilicoevaluationandexperimentaldataconfirmthatthisprimerpairisabletodetectthesamedominanttaxonomicgroups(SupplementaryFigureS2andSupplementaryTableS46).However,incomparisonwiththe16SpyrotagsgeneratedwithS-D-Bact-0341-b-S-17/S-D-Bact-0785-a-A-21andmetagenomestudiesAlphaproteobacteriaappeartobemoreabundantthroughoutallsamples.Bacteroidetes,ontheotherhand,areunder-represented.Asimilarbiascanbefoundonthegrouplevel(SupplementaryFigureS3andSupplementaryTableS47).UseofthisprimerpairindicatesahigherrelativeabundanceofAlphaproteobacteriaSAR11cladesurface1aswellasNAC11-7lineageon7April2009and14April2009.Inturn,particularlythegenusFormosaislessprominent.Thisisinlinewiththeresultsfromtheinsilicoevaluation,whichshowsthatS-D-Bact-0008-a-S-16/S-D-Bact-0907-a-A-20onlydetects52.9%oftheFormosasequences.Evenoneallowedmismatchresultsonlyinanincreaseof9%upto61.9%.Acloserlookrevealsamismatchofthereverseprimertowardsthe3′endforseveralFormosasequences.AlthoughS-D-Bact-0008-a-S-16/S-D-Bact-0907-a-A-20isabletodetectallmajorgroups,abiasintherelativeabundancesaswellascommunitystructureisclearlyconfirmedbytheexperimentaldata(Figure1andSupplementaryFiguresS1–S3).Thissupportsourassumptionthattheoverallcoverageneedalwaystobeconsideredincombinationwiththephylumspectrum.Detailedanalysisofthemismatchpositionshouldalsobetakenintoaccount.Nevertheless,theexperimentalresultsstronglyindicatethatinsilicoevaluationcanserveasaguidelineforchoosingthemostsuitableprimerpair.CONCLUSIONS Theadventofnewsequencingmethodshasbeenaparadigmshiftformolecularecologyandespeciallymicrobialdiversityanalysisusingmarkergenes.Therapidadoptionofthenewtechniquescausedabackloginproperevaluationoftheprimersusedfordiversitysurveys.Ourstudyshowsthatevencommonlyusedsingleprimersexhibitsignificantdifferencesinoverallcoverageandphylumspectrum.Consequently,primerpairsneedtobecarefullyselectedtoavoidaccumulativebias.Outofthe175primersand512primerpairschecked,only10canberecommendedasbroadrangeprimers.Althoughnoneofthemareperfect,andespeciallyfortheArchaeawerecommendthedesignofadditionalprimers,theexperimentalvalidationshowsthatagoodcombinationofprimersapproximatePCR-freemetagenomicapproacheswithrespecttocommunitystructureandrelativeabundances.Theexperimentalresultsconfirmthatsingleinternalmismatches,whenlocatedtowardsthe5′end,aretoleratedintheamplificationprocess.Re-inspectionoftheprimersusingGOSmetagenomeswasfoundtobeareasonableapproachfordeterminingpossibleprimerbiasinthepublicrDNArepositories.However,theincompletephylumspectrumaswellasthecomparativelysmalldatasetsizewithrespectto16SrRNAgenesintheGOSmetagenomesdidnotallowforanin-depthre-evaluation.Forexample,GroupMprimerpairS-D-Bact-0341-b-S-17/S-D-Bact-0785-a-A-21,whichwerecommendedbasedontheSSURef108NRresults,failstodetectmajorgroupsintheGOSdataset,yetexcelsintheexperimentalevaluation.Thisdemonstratesthevalidityofusingcomprehensive,non-redundantdatasetsliketheSILVASSURef108NRfordetailedevaluationofprobesandprimers.WewouldliketonotethattheSILVAprojecthaspreparedanonlineserviceforthispurposeatwww.arb-silva.de/search/testprime,whichismodelledafterourevaluationmethodandallowsinspectionofper-taxoncoveragesforindividualprimerpairs.Furthermore,allprimers,includingbibliographicinformationandinformationonspecificityandoverallcoverage,havebeenaddedtoprobeBase.Theavailabilityoftheevaluatedprimersinacentralandpublicallyaccessiblerepositoryplustheonlineprimerevaluationtoolshouldfacilitatethesearchfor,andtheevaluationandselectionof,suitableprimersinfuturestudies.AVAILABILITY Supplementaryfilesareavailableatwww.arb-silva.de/download/archive/primer_evaluation/.SUPPLEMENTARYDATA SupplementaryDataareavailableatNAROnline:SupplementaryTables1–52andSupplementaryFigures1–3.FUNDING MaxPlanckSocietyandtheFederalMinistryofEducationandResearchGermany[03F0480D].ResearchinthelabofM.H.isfundedthroughgrantsfromtheAustrianResearchFund[Y277-B03];theEuropeanResearchCouncil[StartingGrantEVOCHLAMY281633].Fundingforopenaccesscharge:MaxPlanckSociety.Conflictofintereststatement.Nonedeclared.ACKNOWLEDGEMENTS WeacknowledgeJackA.Gilbert(ArgonneNationalLaboratory,Argonne,IL,USA),BernhardM.Fuchs(MaxPlanckInstitute,Bremen,Germany)andChristineKlockowforcriticaldiscussionofthismanuscript.E.KaramehmedovicandM.Meinersforhelpingwiththelaboratorywork.G.GerdsandA.WichelsfromtheAlfredWegnerInstitute(Bremerhaven,Germany)forsupportingandperformingthewatersampling.HannahMarchant,ElizabethRobertson,MiraOkshevskyandMarioSchimakforcriticalreadingofthemanuscript.REFERENCES 1NossaC.W., OberdorfW.E., YangL., AasJ.A., PasterB.J., DesantisT.Z., BrodieE.L., MalamudD., PolesM.A., PeiZ.. Designof16SrRNAgeneprimersfor454pyrosequencingofthehumanforegutmicrobiome, WJG, 2010, vol. 16 (pg. 4135-4144)GoogleScholarCrossrefSearchADSPubMedWorldCat 2MediniD., SerrutoD., ParkhillJ., RelmanD.A., DonatiC., MoxonR., FalkowS., RappuoliR.. Microbiologyinthepost-genomicera, Nat.Rev.Microbiol., 2008, vol. 6 (pg. 419-430)GoogleScholarCrossrefSearchADSPubMedWorldCat 3MardisE.R.. Theimpactofnext-generationsequencingtechnologyongenetics, TrendsGenet., 2008, vol. 24 (pg. 133-141)GoogleScholarCrossrefSearchADSPubMedWorldCat 4ArmougomF., RaoultD.. Exploringmicrobialdiversityusing16SrRNAhigh-throughputmethods, J.Comput.Sci.Syst.Biol., 2009, vol. 2 (pg. 74-92)GoogleScholarCrossrefSearchADSWorldCat 5MarguliesM., EgholmM., AltmanW.E., AttiyaS., BaderJ.S., BembenL.A., BerkaJ., BravermanM.S., ChenY.J., ChenZ., etal. Genomesequencinginmicrofabricatedhigh-densitypicolitrereactors, Nature, 2005, vol. 437 (pg. 376-380)GoogleScholarCrossrefSearchADSPubMedWorldCat 6SoginM.L., MorrisonH.G., HuberJ.A., MarkWelchD., HuseS.M., NealP.R., ArrietaJ.M., HerndlG.J.. Microbialdiversityinthedeepseaandtheunderexplored‘rarebiosphere’, Proc.NatlAcad.Sci.USA, 2006, vol. 103 (pg. 12115-12120)GoogleScholarCrossrefSearchADSWorldCat 7LiuZ., LozuponeC., HamadyM., BushmanF.D., KnightR.. Shortpyrosequencingreadssufficeforaccuratemicrobialcommunityanalysis, NucleicAcidsRes, 2007, vol. 35 pg. e120 GoogleScholarCrossrefSearchADSPubMedWorldCat 8BennettS.. SolexaLtd, Pharmacogenomics, 2004, vol. 5 (pg. 433-438)GoogleScholarCrossrefSearchADSPubMedWorldCat 9CaporasoJ.G., LauberC.L., WaltersW.A., Berg-LyonsD., HuntleyJ., FiererN., OwensS.M., BetleyJ., FraserL., BauerM., etal. Ultra-high-throughputmicrobialcommunityanalysisontheIlluminaHiSeqandMiSeqplatforms, ISMEJ, 2012, vol. 6 (pg. 1621-1624)GoogleScholarCrossrefSearchADSPubMedWorldCat 10RothbergJ.M., HinzW., RearickT.M., SchultzJ., MileskiW., DaveyM., LeamonJ.H., JohnsonK., MilgrewM.J., EdwardsM., etal. Anintegratedsemiconductordeviceenablingnon-opticalgenomesequencing, Nature, 2011, vol. 475 (pg. 348-352)GoogleScholarCrossrefSearchADSPubMedWorldCat 11EidJ., FehrA., GrayJ., LuongK., LyleJ., OttoG., PelusoP., RankD., BaybayanP., BettmanB., etal. Real-timeDNAsequencingfromsinglepolymerasemolecules, Science, 2009, vol. 323 (pg. 133-138)GoogleScholarCrossrefSearchADSPubMedWorldCat 12LomanN.J., MisraR.V., DallmanT.J., ConstantinidouC., GharbiaS.E., WainJ., PallenM.J.. Performancecomparisonofbenchtophigh-throughputsequencingplatforms, Nat.Biotechnol, 2012, vol. 30 (pg. 434-439)GoogleScholarCrossrefSearchADSPubMedWorldCat 13SchlossP.D., GeversD., WestcottS.L.. ReducingtheeffectsofPCRamplificationandsequencingartifactson16SrRNA-basedstudies, PLoSONE, 2011, vol. 6 pg. e27310 GoogleScholarCrossrefSearchADSPubMedWorldCat 14BakerG.C., SmithJ.J., CowanD.A.. Reviewandre-analysisofdomain-specific16Sprimers, J.Microbiol.Methods, 2003, vol. 55 (pg. 541-555)GoogleScholarCrossrefSearchADSPubMedWorldCat 15WangY., QianP.Y.. Conservativefragmentsinbacterial16SrRNAgenesandprimerdesignfor16SribosomalDNAampliconsinmetagenomicstudies, PLoSONE, 2009, vol. 4 pg. e7401 GoogleScholarCrossrefSearchADSPubMedWorldCat 16TringeS.G., HugenholtzP.. Arenaissanceforthepioneering16SrRNAgene, Curr.Opin.Microbiol., 2008, vol. 11 (pg. 442-446)GoogleScholarCrossrefSearchADSPubMedWorldCat 17HamadyM., KnightR.. Microbialcommunityprofilingforhumanmicrobiomeprojects:tools,techniques,andchallenges, Genome.Res., 2009, vol. 19 (pg. 1141-1152)GoogleScholarCrossrefSearchADSPubMedWorldCat 18AnderssonA.F., LindbergM., JakobssonH., BäckhedF., NyrénP., ngstrandL.. Comparativeanalysisofhumangutmicrobiotabybarcodedpyrosequencing, PLoSONE, 2008, vol. 3 pg. e2836 GoogleScholarCrossrefSearchADSPubMedWorldCat 19LiuZ., DeSantisT.Z., AndersenG.L., KnightR.. Accuratetaxonomyassignmentsfrom16SrRNAsequencesproducedbyhighlyparallelpyrosequencers, NucleicAcidsRes., 2008, vol. 36 pg. e120 GoogleScholarCrossrefSearchADSPubMedWorldCat 20PruesseE., QuastC., KnittelK., FuchsB.M., LudwigW., PepliesJ., GlöcknerF.O.. SILVA:acomprehensiveonlineresourceforqualitycheckedandalignedribosomalRNAsequencedatacompatiblewithARB, NucleicAcidsRes., 2007, vol. 35 (pg. 7188-7196)GoogleScholarCrossrefSearchADSPubMedWorldCat 21AlmE.W., OertherD.B., LarsenN., StahlD.A., RaskinL.. Theoligonucleotideprobedatabase, Appl.Environ.Microbiol., 1996, vol. 62 (pg. 3557-3559)GoogleScholarPubMedOpenURLPlaceholderTextWorldCat 22TeelingH., FuchsB.M., BecherD., KlockowC., GardebrechtA., BennkeC.M., KassabgyM., HuangS., MannA.J., WaldmannJ., etal. Substrate-controlledsuccessionofmarinebacterioplanktonpopulationsinducedbyaphytoplanktonbloom, Science, 2012, vol. 336 (pg. 608-611)GoogleScholarCrossrefSearchADSPubMedWorldCat 23BrosiusJ., PalmerM.L., KennedyP.J., NollerH.F.. Completenucleotidesequenceofa16SribosomalRNAgenefromEscherichiacoli, Proc.NatlAcad.Sci.USA, 1978, vol. 75 (pg. 4801-4805)GoogleScholarCrossrefSearchADSWorldCat 24LoyA., MaixnerF., WagnerM., HornM.. probeBase–anonlineresourceforrRNA-targetedoligonucleotideprobes:newfeatures2007, NucleicAcidsRes., 2007, vol. 35 (pg. D800-D804)GoogleScholarCrossrefSearchADSPubMedWorldCat 25KibbeW.A.. OligoCalc:anonlineoligonucleotidepropertiescalculator, NucleicAcidsRes., 2007, vol. 35 (pg. W43-W46)GoogleScholarCrossrefSearchADSPubMedWorldCat 26PruesseE., PepliesJ., GlöcknerF.O.. SINA:accuratehighthroughputmultiplesequencealignmentofribosomalRNAgenes, Bioinformatics, 2012, vol. 28 (pg. 1823-1829)GoogleScholarCrossrefSearchADSPubMedWorldCat 27EdgarR.C.. SearchandclusteringordersofmagnitudefasterthanBLAST, Bioinformatics, 2010, vol. 26 (pg. 2460-2461)GoogleScholarCrossrefSearchADSPubMedWorldCat 28RuschD.B., HalpernA.L., SuttonG., HeidelbergK.B., WilliamsonS., YoosephS., WuD., EisenJ.A., HoffmanJ.M., RemingtonK., etal. TheSorcererIIGlobalOceanSamplingexpedition:northwestAtlanticthrougheasterntropicalPacific, PLoSONE, 2007, vol. 5 pg. e77 GoogleScholarCrossrefSearchADSWorldCat 29YoosephS., SuttonG., RuschD.B., HalpernA.L., WilliamsonS.J., RemingtonK., EisenJ.A., HeidelbergK.B., ManningG., LiW., etal. TheSorcererIIGlobalOceanSamplingexpedition:expandingtheuniverseofproteinfamilies, PLoSBiol., 2007, vol. 5 pg. e16 GoogleScholarCrossrefSearchADSPubMedWorldCat 30LudwigW., StrunkO., WestramR., RichterL., MeierH., KumarY., BuchnerA., LaiT., SteppiS., JobbG., etal. ARB:asoftwareenvironmentforsequencedata, NucleicAcidsRes., 2004, vol. 32 (pg. 1363-1371)GoogleScholarCrossrefSearchADSPubMedWorldCat 31ZhouJ., BrunsM.A., TiedjeJ.M.. DNArecoveryfromsoilsofdiversecomposition, Appl.Environ.Microbiol., 1996, vol. 62 (pg. 316-322)GoogleScholarPubMedOpenURLPlaceholderTextWorldCat 32HerlemannD.P.R., LabrenzM., JuergensK., BertilssonS., WaniekJ.J., AnderrsonA.F.. Transitioninbacterialcommunitiesalongthe2000kmsalinitygradientoftheBalticSea, ISMEJ, 2011, vol. 5 (pg. 1571-1579)GoogleScholarCrossrefSearchADSPubMedWorldCat 33MuyzerG., deWaalE.C., UitterlindenA.G.. Profilingofcomplexmicrobialpopulationsbydenaturinggradientgelelectrophoresisanalysisofpolymerasechainreaction-amplifiedgenescodingfor16SrRNA, Appl.Environ.Microbiol., 1993, vol. 59 (pg. 695-700)GoogleScholarPubMedOpenURLPlaceholderTextWorldCat 34MuyzerG., BrinkhoffT., NuebelU., SantegoedsC., SchaeferH., WaverC.. AkkermansA.D.L., vanElsasJ.D., deBruijnF.J.. Denaturinggradientgelelectrophoresis(DGGE)inmicrobialecology, MolecularMicrobialEcologyManual, 1998Dordrecht,TheNetherlandsKluwerAcademicPublishers(pg. 1-27)GoogleScholarGooglePreviewOpenURLPlaceholderTextWorldCatCOPAC 35GilbertJ.A., FieldD., SwiftP., ThomasS., CummingsD., TempertonB., WeynbergK., HuseS., HughesM., JointI., etal. Thetaxonomicandfunctionaldiversityofmicrobesatatemperatecoastalsite:a'multi-omic'studyofseasonalanddieltemporalvariation, PLoSONE, 2010, vol. 5 pg. e15545 GoogleScholarCrossrefSearchADSPubMedWorldCat 36WaltersW.A., CaporasoJ.G., LauberC.L., Berg-LyonsD., FiererN., KnightR.. PrimerProspector:denovodesignandtaxonomicanalysisofbarcodedPCRprimers, Bioinformatics, 2011, vol. 27 (pg. 1159-1161)GoogleScholarCrossrefSearchADSPubMedWorldCat 37BarnsS.M., FundygaR.E., JeffriesM.W., PaceN.R.. RemarkablearchaealdiversitydetectedinaYellowstoneNationalParkhotspringenvironment, Proc.NatlAcad.Sci.USA, 1994, vol. 91 (pg. 1609-1613)GoogleScholarCrossrefSearchADSWorldCat 38ReysenbachA.L., PaceN.R.. RobbF.T., PlaceA.R.. , Archaea:ALaboratoryManual—Thermophiles.CSHLP.Protocol16, 1995(pg. 101-107) 39HuwsS.A., EdwardsJ.E., KimE.J., ScollanN.D.. Specificityandsensitivityofeubacterialprimersutilizedformolecularprofilingofbacteriawithincomplexmicrobialecosystems, J.Microbiol.Methods, 2007, vol. 70 (pg. 565-569)GoogleScholarCrossrefSearchADSPubMedWorldCat 40DroegeM., HillB.. ThegenomesequencerFLXsystem–longerreads,moreapplications,straightforwardbioinformaticsandmorecompletedatasets, J.Biotechnol, 2008, vol. 136 (pg. 3-10)GoogleScholarCrossrefSearchADSPubMedWorldCat 41SiposR., SzékelyA.J., PalatinszkyM., RévészS., MárialigetiK., NikolauszM.. Effectofprimermismatch,annealingtemperatureandPCRcyclenumberon16SrRNAgene-targettingbacterialcommunityanalysis, FEMSMicrobiol.Ecol., 2007, vol. 60 (pg. 341-350)GoogleScholarCrossrefSearchADSPubMedWorldCat 42BakerG.C., CowanD.A.. 16SrDNAprimersandtheunbiasedassessmentofthermophilediversity, Biochem.Soc.Trans., 2004, vol. 32 (pg. 218-221)GoogleScholarCrossrefSearchADSPubMedWorldCat 43HuberH., HohnM.J., RachelR., FuchsT., WimmerV.C., StetterK.O.. AnewphylumofArchaearepresentedbyananosizedhyperthermophilicsymbiont, Nature, 2002, vol. 417 (pg. 63-67)GoogleScholarCrossrefSearchADSPubMedWorldCat 44HuseS.M., DethlefsenL., HuberJ.A., MarkWelchD., RelmanD.A., SoginM.L.. ExploringmicrobialdiversityandtaxonomyusingSSUrRNAhypervariabletagsequencing, PLoSGenet., 2008, vol. 4 pg. e1000255 GoogleScholarCrossrefSearchADSPubMedWorldCat 45McClimentE.A., NelsonC.E., CarlsonC.A., AlldredgeA.L., WittingJ., Amaral-ZettlerL.A.. Anall-taxonmicrobialinventoryoftheMooreacoralreefecosystem, ISMEJ., 2012, vol. 6 (pg. 309-319)GoogleScholarCrossrefSearchADSPubMedWorldCat 46AbedR.M., Al-ThukairA., deBeerD.. Bacterialdiversityofacyanobacterialmatdegradingpetroleumcompoundsatelevatedsalinitiesandtemperatures, FEMSMicrobiol.Ecol., 2006, vol. 57 (pg. 290-301)GoogleScholarCrossrefSearchADSPubMedWorldCat 47LiebnerS., HarderJ., WagnerD.. BacterialdiversityandcommunitystructureinpolygonaltundrasoilsfromSamoylovIsland,LenaDelta,Siberia, Int.Microbiol., 2008, vol. 11 (pg. 195-202)GoogleScholarPubMedOpenURLPlaceholderTextWorldCat 48SchauerR., BienholdC., RametteA., HarderJ.. Bacterialdiversityandbiogeographyindeep-seasurfacesedimentsoftheSouthAtlanticOcean, ISMEJ., 2010, vol. 4 (pg. 159-170)GoogleScholarCrossrefSearchADSPubMedWorldCat 49DeSantisT.Z., HugenholtzP., LarsenN., RojasM., BrodieE.L., KellerK., HuberT., DaleviD., HuP., AndersenG.L.. Greengenes,achimera-checked16SrRNAgenedatabaseandworkbenchcompatiblewithARB, Appl.Environ.Microbiol., 2006, vol. 72 (pg. 5069-5072)GoogleScholarCrossrefSearchADSPubMedWorldCat 50ColeJ.R., ChaiB., FarrisR.J., WangQ., Kulam-Syed-MohideenA.S., McGarrellD.M., BandelaA.M., CardenasE., GarrityG.M., TiedjeJ.M.. Theribosomaldatabaseproject(RDP-II):introducingmyRDPspaceandqualitycontrolledpublicdata, NucleicAcidsRes., 2007, vol. 35 (pg. D169-D172)GoogleScholarCrossrefSearchADSPubMedWorldCat 51YoosephS., NealsonK.H., RuschD.B., McCrowJ.P., DupontC.L., KimM., JohnsonJ., MontgomeryR., FerrieraS., BeesonK., etal. Genomicandfunctionaladaptationinsurfaceoceanplanktonicprokaryotes, Nature, 2010, vol. 468 (pg. 60-66)GoogleScholarCrossrefSearchADSPubMedWorldCat 52GilbertJ.A., MeyerF., JanssonJ., GordonJ., PaceN., TiedjeJ., LeyR., FiererN., FieldD., KyrpidesN., etal. TheEarthMicrobiomeProject:meetingreportofthe“1stEMPmeetingonsampleselectionandacquisition”atArgonneNationalLaboratoryOctober6th2010, Stand.GenomicSci., 2010, vol. 3 (pg. 249-253)GoogleScholarCrossrefSearchADSPubMedWorldCat  ©TheAuthor(s)2012.PublishedbyOxfordUniversityPress.ThisisanOpenAccessarticledistributedunderthetermsoftheCreativeCommonsAttributionLicense(http://creativecommons.org/licenses/by-nc/3.0/),whichpermitsnon-commercialreuse,distribution,andreproductioninanymedium,providedtheoriginalworkisproperlycited.Forcommercialre-use,[email protected].©TheAuthor(s)2012.PublishedbyOxfordUniversityPress. Subject Nucleicacidamplification IssueSection: MethodsOnline Downloadallslides Supplementarydata Supplementarydata AddSuppFiles-6-zipfile AddSuppFiles-5-zipfile AddSuppFiles-4-zipfile AddSuppFiles-3-zipfile AddSuppFiles-2-zipfile AddSuppFiles-1-zipfile Comments 0Comments Comments(0) Addcomment Closecommentformmodal Iagreetothetermsandconditions. Youmustacceptthetermsandconditions. Addcomment Cancel Submitacomment Name Affiliations Commenttitle Comment Youhaveenteredaninvalidcode Submit Cancel Thankyouforsubmittingacommentonthisarticle.Yourcommentwillbereviewedandpublishedatthejournal'sdiscretion.Pleasecheckforfurthernotificationsbyemail. Advertisement Citations Views 52,141 Altmetric Moremetricsinformation × Emailalerts Articleactivityalert Advancearticlealerts Newissuealert Subjectalert ReceiveexclusiveoffersandupdatesfromOxfordAcademic Moreonthistopic IdentificationandclassificationofantiviraldefencesystemsinbacteriaandarchaeawithPADLOCrevealsnewsystemtypes InsightsintosynthesisandfunctionofKsgA/Dim1-dependentrRNAmodificationsinarchaea BypassingrRNAmethylationbyRsmA/Dim1duringribosomematurationinthehyperthermophilicarchaeonNanoarchaeumequitans CrystalstructureofPyrococcushorikoshiitryptophanyl-tRNAsynthetaseandstructure-basedphylogeneticanalysissuggestanarchaealoriginoftryptophanyl-tRNAsynthetase Relatedarticlesin WebofScience GoogleScholar RelatedarticlesinPubMed StandardisationofACPAtests:evaluationofanewcandidatereferencepreparation. IL-10andTGF-βIncreaseConnexin-43ExpressionandMembranePotentialofHL-1CardiomyocytesCoupledwithRAW264.7Macrophages. PROcalcitoninandNEWS2evaluationforTimelyidentificationofsepsisandOptimaluseofantibioticsintheemergencydepartment(PRONTO):protocolforamulticentre,open-label,randomisedcontrolledtrial. 6α-hydroxylatedbileacidsmediateTGR5signallingtoimproveglucosemetabolismupondietaryfibersupplementationinmice. Citingarticlesvia WebofScience(3462) GoogleScholar Crossref Latest MostRead MostCited Quality-controlledR-loopmeta-analysisrevealsthecharacteristicsofR-loopconsensusregions RUV-III-NB:normalizationofsinglecellRNA-seqdata GuidingthefoldingofG-quadruplexesthroughloopresidueinteractions Selectiveablationof3′RNAendsandprocessiveRTsfacilitatedirectcDNAsequencingoffull-lengthhostcellandviraltranscripts StructuralanddynamicstudiesofDNArecognitionbyNF-κBp50RHRhomodimerusingmethylNMRspectroscopy Advertisement Advertisement AboutNucleicAcidsResearch EditorialBoard Policies AuthorGuidelines Facebook Twitter LinkedIn AdvertisingandCorporateServices JournalsCareerNetwork OnlineISSN1362-4962PrintISSN0305-1048Copyright©2022OxfordUniversityPress AboutUs ContactUs Careers Help Access&Purchase Rights&Permissions OpenAccess PotentiallyOffensiveContent Connect JoinOurMailingList OUPblog Twitter Facebook YouTube LinkedIn Resources Authors Librarians Societies Sponsors&Advertisers Press&Media Agents Explore ShopOUPAcademic OxfordDictionaries Epigeum OUPWorldwide UniversityofOxford OxfordUniversityPressisadepartmentoftheUniversityofOxford.ItfurtherstheUniversity'sobjectiveofexcellenceinresearch,scholarship,andeducationbypublishingworldwide Copyright©2022OxfordUniversityPress CookiePolicy PrivacyPolicy LegalNotice SiteMap Accessibility Close ThisFeatureIsAvailableToSubscribersOnly SignInorCreateanAccount Close ThisPDFisavailabletoSubscribersOnly ViewArticleAbstract&PurchaseOptions Forfullaccesstothispdf,signintoanexistingaccount,orpurchaseanannualsubscription. Close