Robust species taxonomy assignment algorithm for 16S rRNA NGS reads: application to oral carcinoma samples

Journal of Oral Microbiology

Akram Thabet Nasher¹  Tsute Chen²  Nezar Noor Al-hebshi³  Ali Mohamed Idris⁴ 

[1] Sana'a University;Forsyth Institute;Jazan University;Jazan Univeristy

关键词: bacteria;    cancer;    carcinogenesis;    next generation sequencing;    OSCC;    pyrosequencing;   

DOI  :  

来源: Co-Action Publishing

【 摘 要 】

Background:Usefulnessofnext-generationsequencing(NGS)inassessingbacteriaassociatedwithoralsquamouscellcarcinoma(OSCC)hasbeenunderminedbyinabilitytoclassifyreadstothespecieslevel.Objective:Thepurposeofthisstudywastodeveloparobustalgorithmforspecies-levelclassificationofNGSreadsfromoralsamplesandtopilottestitforprofilingbacteriawithinOSCCtissues.Methods:Bacterial16SV1-V3librarieswerepreparedfromthreeOSCCDNAsamplesandsequencedusing454’sFLXchemistry.High-quality,well-aligned,andnon-chimericreads≥350bpwereclassifiedusinganovel,multi-stagealgorithmthatinvolvesmatchingreadstoreferencesequencesinrevisedversionsoftheHumanOralMicrobiomeDatabase(HOMD),HOMDextended(HOMDEXT),andGreengeneGold(GGG)atalignmentcoverageandpercentageidentity≥98%,followedbyassignmenttospecieslevelbasedontophitreferencesequences.PrioritywasgiventohitsinHOMD,thenHOMDEXTandfinallyGGG.Unmatchedreadsweresubjecttooperationaltaxonomicunitanalysis.Results:Nearly,92.8%ofthereadswerematchedtoupdated-HOMD13.2,1.83%totrusted-HOMDEXT,and1.36%tomodified-GGG.Ofallmatchedreads,99.6%wereclassifiedtospecieslevel.Atotalof228species-leveltaxawereidentified,representing11phyla;themostabundantwereProteobacteria,Bacteroidetes,Firmicutes,Fusobacteria,andActinobacteria.Thirty-fivespecies-leveltaxaweredetectedinallsamples.Onaverage,Prevotellaoris,Neisseriaflava,Neisseriaflavescens/subflava,Fusobacteriumnucleatumsspolymorphum,Aggregatibactersegnis,Streptococcusmitis,andFusobacteriumperiodontiumwerethemostabundant.Bacteroidesfragilis,aspeciesrarelyisolatedfromtheoralcavity,wasdetectedintwosamples.Conclusion:Thismulti-stagealgorithmmaximizesthefractionofreadsclassifiedtothespecieslevelwhileensuringreliableclassificationbygivingprioritytothehuman,oralreferenceset.ApplyingthealgorithmtoOSCCsamplesrevealedhighdiversity.Inadditiontooraltaxa,anumberofhuman,non-oraltaxawerealsoidentified,someofwhicharerarelydetectedintheoralcavity.Keywords:bacteria;cancer;next-generationsequencing;OSCC;pyrosequencing;taxonomy(Published:29September2015)Citation:JournalofOralMicrobiology2015,7:28934-http://dx.doi.org/10.3402/jom.v7.28934

【 授权许可】

Unknown