Quick viewing(Text Mode)

Improved Accuracy from Full-Length 16S Multiplexed Strain Identification Mock Community Classification

Improved Accuracy from Full-Length 16S Multiplexed Strain Identification Mock Community Classification

Accurately Surveying Uncultured Microbial with SMRT® Sequencing B. Bowman1, J-F. Cheng2, E. Singer2, D. Coleman-Derr2, S. Tringe2, S. Hallam3, T. Woyke2 1Pacific Biosciences, 1380 Willow Road, Menlo Park, CA 94025, 2Dept. of Energy Joint Inst., Walnut Creek, CA, 3Univ. of British Columbia, Vancouver, BC, Canada

Introduction Improved Accuracy from Full-Length 16S Multiplexed Strain Identification Mock Community Classification

Analysis of bacterial Mock Pilot project for multiplexed Species OTU Count Concordance V4 Full Microbial ecology is reshaping our understanding of the natural world by revealing the large w/ Reference Class. Class. Community, composed of a strain identification from phylogenetic and functional diversity of microbial . However the vast majority of these A perfringens 1 99.92% Species mixture of 22 bacterial species barcoded full-length 16S remain poorly understood, as most cultivated representatives belong to just four Clostridium thermocellum 1 99.86% Genus Strain for which reference assemblies sequence data using phylogenetic groups and more than half of all identified phyla remain uncultivated. Characterization Coraliomargarita was available. Full-length 16S asymmetric, paired barcodes akajimensis 1 100% Genus Strain of this microbial ‘dark matter’ will thus greatly benefit from new metagenomic methods for in situ amplicons were generated with analysis. For example, sensitive high throughput methods for the characterization of community and Long Amplicon Analysis. glutamicum 1 99.79% Genus Strain the 27F/1492R primer pair and composition and structure from the sequencing of conserved marker . Desulfosporosinus sequenced on the PacBio RS II. Figure 6 (Left-Top): acidiphilus 1 100% Genus Strain Sequence data was analyzed Here we utilize Single Molecule Real-Time (SMRT®) sequencing of full-length 16S rRNA amplicons Histogram of post-filter Desulfosporosinus meridiei 1 99.67% Genus Strain with rDnaTools[3]. B Desulfotomaculum to phylogenetically profile microbial communities to below the genus-level. We test this method with subread counts by barcode. gibsoniae 5 98.2%-100% Family Strain three different data sets: a mock community of known composition, a previously studied microbial Unlabeled subreads account vietnamensis 1 100% Genus Strain Table 2 (Left): community from a lake known to predominantly contain poorly characterized phyla[1], and a for < 2.5% of the over 200,000 Escherichia coli 1 99.86% Family Species Analysis of the OTU consensus multiplexed. These results are compared to traditional 16S tag sequencing of the V4 hyper-variable high-quality subreads. The Fervidobacterium sequences generated by region with short-read sequencing technologies. We explore the benefits of using full-length dashed line represents the pennivorans 1 99.93% Genus Strain rDnaTools. Classification of the ribosomal DNA amplicons for estimating community structure and diversity, as well as for single- 500 subreads recommended Frateuria aurantia 1 99.93% Genus Strain Figure 4. Hirschia baltica 1 100% Genus Strain V4 region was carried out with isolate strain identification relative to alternate methods. We characterize the potential benefits of for phasing. A) Comparison of differences between two spp. in the full-length 16S (2.6%) and silvanus 1 100% Genus Species the RDP Classifier[5] on V4 profiling metagenomic communities with full-length 16S rRNA genes from SMRT sequencing relative hyper-variable V4-region (0%) Olsenella uli 1 100% Genus Strain sequences extracted from the to standard methods. Pseudomonas stutzeri 1 100% Genus Strain Figure 7 (Left-Bottom): reference . B) Example of the over-estimation of diversity that can result from classification based on small Salmonella bongori 2 99.93% Family Strain hyper-variability regions compared to the full-length sequence Histogram of consensus 1 99.24% Genus Serovar Classification of full-length 16S sequence counts by barcode. rotundus 1 99.93% Genus Strain was performed with RDP Multiple consensus sequences smaragdinae 1 100% Genus Strain SeqMatch[5] on the OTU pyogenes 1 100% Genus Strain Reads-of-Insert Consensus enable better classification, consensus sequences Environmental Sample Terriglobus roseus 1 100% Genus Strain 0.1 generated by rDnaTools Thermobacillus composti 1 100% Genus Strain Community analysis of environmental Reads-of-Insert (RoI) Consensus Sequences 16S sequence data from Lake Sakinaw,

B

B

B a

B a

B a c a c B c t

a t e B c B t e a c e r ; t a B e r a c t i r a i 3 e c ; a t r i a c ; a ; e t r i F 9 ; C c t a e I i 9 e r ; C a F B t C i ; r h I F e r a C G i ; h I a a i ; C h l r a ; G h l o

i a c C o ; a l ; G ; ; i h C o r ; t C l r a 9 e ; h o o l d i r o a C o h i i h r o r l f F o d o f l I i r l o i h e d l f l o a o i o r l e ; f c l o e x G ; o l r o r f o x ; 9 e c C o c l ; o f x i e c r i ; a o l x c F ; f 9 h o f e i D i c I x l ; D ; l i c e o D l f e ; d F x i e o i o D G I 9 l o ; e c x e x i l ; ; D e c r ; h o F o x i D e h i o G 9 a a I ; l ; h o e a c i ; D l ; f i D h a h a F ; e a c l h l d a G a i I 9 e D l o i a e ; e e h h l o o ; a o h d F x l c o a G h e h e c i I a o c i ; 9 l D i o e c o c a ; ; ; h a o o o l c d F i a D l G D o o c i c i I c ; a l ; D c 9 l c o x i o o ; a o o c c c d e o i c a ; c F l c i G e x i o c c c I c o c o h ; h c l x o 9 c o o f e d o c PacBio’s long read-length and circularized c o c o e o a i c i a c l e l i G F ; o i d c o f l o ; I o c d c c o i o l f D o c 9 British Columbia. Illumina V4-region r B d a d i i U l o U c a c o ; i ; c c U o i a c B i d i i G ; o o F ; a h a d i o r a ; d c o o o I c o a l ; c i ; r i n a n ; x l c d n d 9 c d a ; e v o i e o o i h c a i i d ; h l o a v o i r c c G d d c d e a o F t i c c ; v a e l l e l c o ; ; e i a ; D c i I t a e h d u v l d ; l f h i C e ; e a i l f i h ; c i t a 9 d o c v a a a i i r a a a l d ; i ; f o a d D i Conclusion G i f ; o e l r i v i C c o ; d d c o s F a i s x ; h ; a a i u ; s ; ; s ; C e r i a i d I d i n ; s d a a v i i o c 2 a F ; i n o c i i s e D ; ; d s s e e a s s f o l 9 d n r a i x ; i C i B d l c ; U v a s i e l h c o G 0 n d i i a i n B W f f C i s i i i F a i i a ; i f e n B f e r ; D d ; a f s e o o c l f e I − u h U d a h l x a A i i ; i n i t a r i o ; B l A i l i h n e s f h a e i c 9 e l e s f d c ; d B ; i c r l i 2 e B ; A i o U c e D a G e 2 t s l o n c n c C A s o l x e N B d d 2 d c t e d ; o o d ; F o d a i n ; o f c 3 A 2 s i r i l a a i I r l n 2 c l s l h c B 6 n n a e c e a ; f a r o ; n c c 2 A 2 6 e D a ; o l l s a i o x a o i e r ; Unclassified; c e o l B 6 U a a i h f ; 2 ; i B r G c A l l i o o f a s U − a c U p 6 ; f u 2 c a r e h l c d ; f l U n i t ; l 6 ; o l i l e s s A B c l C h o x D c 7 a l o a 2 ; n e ss l r f ; e a c a e U n 6 B s ; o i d x ; c i e u s i U t e o 1 x l s f 2 6 n C h o o l h l o e c ; a s c c i i U a l x D d n r U n i i n c l i ; r f c i 1 i ; − s e 6 ; e a c i n ; D l f ; U a r C e i l u D c a i a c l a h o o l o o S i t 4 i U i ; l h u U n f e d ; r f x D l o ; l e l s n c e r c F d e a B i B t a C h ; e c l u c e d ; i ; o o e i a c ; e h n s n e l l c h l s U c r r x o a U − d ; t C f D h l o a a a c a i a h ; i n a s f U a e i ; o o e C e i 6 ; c l i e a c ; l n l s t r r l d u ; B l o U a i e a C f x i c ; o c s f 5 B a c e i ; h D h lo s d c i d t r o o ; o a e a c n l s i e l e i e e e o U a f ; B a a C r l a c l e a r c o c s ; c e i ; h f x i d t r o D h c a n e u t c c n l s i e B a l o ; li t e f ; c a C e i e o e c l ; c c U a s i d e i r l r c B t ; h f c n o a u d i o o s i e a r o x i r a n l f ; c a l o D o l e c e i a s i d e i C r le i; l i e n r ; id d U c e B t r ; h n C s i ; a f x d a li u i l f d c a o a l ;u lt ia a n a s i B e i C l o n o h e a t ; r e h a r e u ; c s i ; r h l A ; lo ; v f d c o f e C ; e a c d S l s i B e ia l ; r a a e ; a t C o s a e n sequence templates enable the generation of h U i r ; r D e o s f d c h ; e n c u B 7 d i B e a o i a l a ; data was analyzed with iTagger[2], while f n s e ; a t i C l x a A e le 6 in r ; e ; e lture a c i d c h e a c x 5 B f B e a le n s n l ie ; a t i ; f i n li e cu te i; B a r C il li le c n D s A d B c ; d o o r U − e a r d o a r a ;u i e s 2 ; a t i ie o l r e e e a n A r f l e ; h if 6 B c i a n a in C a U c G i ; a e h a li n l h l e t s ;C n o o eae lo n a − d B c s C ; i r A r lo s ; d x A ; e ac c U c 1 ; a la ; e s r o l s ia e le e a e a e o n a i 1 B c r fi ; f l n n f c s f 1 n i a n a l c c ie e d o e A oli e o U l s ; t s e r ;A e ; r x a i d U c s i o in n s e i i n s f a if l l e li e ;D d c ie ; la h o a l ia U l s B c s r e ro a na B e a i d s C e e a h ;N n s fi ; n a ; ; in e n a cl e l a d a l a i c a s d U c i e n o n l es;A te lo a if r i r ro Ba p ss i ; n e if ; ;A e ;A al ri c o e U t s d i a e e e a o U l d c e x a ct ; c i− if ; a s i e n a n lin C c n ie a if ; l A e eria h o c 4 B l s d f i; n ;A ro l i U l B d c s e o x li e e ; o d a − ; n i r o 6 ;C r ia n s la if lo le r a a L o c s 6 U c s f e e n h fl ;S U l i 5 s h o a in A B lo e a fi ; n C r l S x h n s e U la ; o n o e; rofle i 7 c s d c a l A r ;M ;D 6 l i ; ri h i; e ea a 5 a fi n C x a ii e s e U te ; ; n lin r xi h B s d c a le o ; ;De a − i ; i d f ;A 6 high-quality consensus sequences from l U A f a r e o i cte L o ie B e fi r x aer a B PacBio full-length 16S data was c n G d t i lo e b hal c s l S B o c − ; a s h f a c la 1 o so ;M c oc c s B la ;C r u a t o s 1 c a lo exi;An F ii e o id i 1 n ri h l r ri c i fi ; ia; te a c a e U te ;C rof r c ;C oid ;G d c a e a h ; a ri lo b lo ia;G IF B e act o r Unclass 3 t b s o c a;Ch o u fle ; a i ;F x IF3 B er us a i U t ;F ri ;D n ; c e e c ifi a ia ; ct h la ed; B ter a al U s c ified b o n s a s o • The PacBio RSII enables high-throughput sequencing of full- c c if s oc la ie B u c U s d clas ;F o n s ; n a id c if ri ; i la ie U e d a; s d ct fie v s ; a si ad if B s ; i ie la d nB d c fie U A ; n si ; nc 2 U s ; 1 la 6 la ed 1 U s ; c ifi P n si n s O multiple passes over the same molecule. c fie U s d; on l la i analyzed with rDnaTools[3]. as d; c fie is U s n si iv n ifi U s ; d cl ed la d te as ; c fie a U s n si id n ifi U s d cl ed la an a ; c C U ss n ; n ifi U ria cl ed te d; B as ; c ie ac U s a if te n ifi B ss ria c ed a d; ;C la ; cl ie Figure 8 (Below): s n h U s U sif lor n ifie s of cl d a d; lex as ; ncl ie i;D s U sif e Un ifie s ha cl d cla d; lo as ; n ie co si U sif cc Un fie s oi c d; cla ; di las n ied a;N s U if a ifie ss p d la ; oli ; nc d −4 U ifie B− ss 65 cla ; ; Un ied sif length 16S amplicons las nc d; U fie Forward #1 si las nc d; U fie ssi cla n d; U ifie ss B cla acte Un d; ria;P ifie rote ass oba cl Proposed experimental design for the efficient multiplexed n B cter U d; acte ia;D ifie ria; elta ass Prot pro ncl eob teob U d; acte ac ifie Ba ria; teria ss cter Delt ;De cla ia;Pr apro sulf Un ; oteo teo arcu fied bac bac lale ssi teri teria s;D cla Bac a;De ;De esu Un teria ltap sulf lfar d; ;Pro rote arcu cula sifie teob obac lale cea las acte teria s;De e;u Unc B ria;D ;De sul ncu d; acte elta sulf farc lture ifie ria;P prote arcu ulac d; ass rote oba lales eae ncl obac cter ;De ;unc U teria ia;De sulf ultu ed; Figure 5 (Right): Phylogenetic tree of all B ;Delt sulf arcu red ssifi acte apro arcu lace ; cla ria;Pr teob lales ae;u Un oteoba acter ;Des ncu d; cte ia;De ulfar lture sifie ria;De sulfa cula d; clas Ba ltapro rcula ceae Un cteria; teoba les;D ;unc ; Prote cter esu ultur ified screening of 16S amplicons. Current throughput of o ia lfa e s bacter ;Des rcula d; clas ia;De ulfarc ceae Un B ltapro ulale ;unc ; acter teoba s;De ultu ified ia;Pro cteria sulfar red; lass teobac ;Desu culac Unc Ba teria;D lfarcu eae;u cteria;Pr eltapr lales ncult ied; oteob oteob ;Desu ured; lassif acteria acteria lfarcu Unc ;Deltap ;Desu laceae Forward #2 r lfa ; • Full-length 16S amplicons show both greater sensitivity and oteo rcu unc ; bacter lales; ulture sified Bac ia;Des Desulf d; nclas teria;Pro ulfobac arcula U teobac terales ceae;u teria;De ;Desul ncultu ified; ltaprote fobacte red; class obacter raceae Un Bac ia;Synt ;Desul teria;Pro rophoba fatirhab ed; teobact cterales dium; lassifi OTUs (359 assigned) identified from full- eria;Delta ;Syntro Unc proteo phace ; ae;Desu Bacte Syntroph lfomon ssified; ria;Proteoba obacter ile; Uncla cteria;Delta ales;Syn proteobact trophace eria;Syntro ae;uncu ified; phobactera ltured; Unclass >200,000 post-filter subreads with a recommended 500 Bac les;S teria;Proteoba yntrophacea cteria;Deltapr e;unculture oteobacteria; d; ified; Syntrophobac Unclass terales;Syntro phaceae;unc ultured; d; Unclassifie U nclassified; Unclassified;

Unclassified; Unclassified; length PacBio 16S sequences[4]. Pink: Unclassified; Unclassified; Bacteria;;De ltaproteobacteria;;Syntrophaceae;; Unclassified; Assaying 16S Reads-of-Insert Bacteria;Proteobacteria;;Syntrophobacterales;Syntrophaceae;Desulfobacca; Unclassified; subreads per barcode for phasing implies a maximum specificity to the clustering of OTUs than 16S tag sequencing xiellaceae;Aquicella; Bacteria;Proteobacteria;;;Co Unclassified; Unclassified; Unclassified; Forward #3 Chemistry: P4 / C2 Unclassified; Unclassified; egionella; Unclas Legionellaceae;L sified; ria;Legionellales; mmaproteobacte roteobacteria;Ga lla; U Bacteria;P ceae;Legione nclassified; les;Legionella 175 (48.74%) unclassified OTUs; ria;Legionella lla; oteobacte almone ria;Gammapr riaceae;S Unclas a;Proteobacte terobacte sified; Bacteri eriales;En nterobact onas; bacteria;E sychrom Unc maproteo daceae;P lassified; teria;Gam hromona throughput of ~400x samples and an estimated c c ;Proteoba ales;Psy ; Sequence Quality Bacteria omonad occales Un Movie: 1x 120min ria;Alter ethyloc classifie acte ia;M d; aproteob obacter ia;Gamm maprote rium; eobacter ia;Gam ibacte Uncla ria;Prot obacter ae;Und ssified; Bacte ia;Prote terace Bacter xalobac e; riales;O adacea Uncl kholde amon assified ria;Bur s;Com ; obacte deriale a; taprote urkhol tener Unc eria;Be eria;B ethylo lassifie eobact obact eae;M d; ria;Prot taprote philac ; Bacte ria;Be ethylo teuria Ba obacte les;M e;Fra cteria; ;Prote ophila cea Cand acteria ethyl onada idate d Orange: candidate phyla; Blue: phyla B teria;M nthom chia; Un ivisio bac ;Xa irs cla n O System: PacBio RS II roteo dales eae;H ssified D1; etap ona dac ; eria;B hom ona obact ;Xant phom M6; Unc • Comparisons of short-read and full-length 16S sequences te ia y ;T las ria;Pro bacter ales;H teria sifie Bacte oteo cter Bac d; mapr uloba ; U throughput of ~150-200x per SMRTcell. m a ia;Ga ria;C ;TM6 ncla acter acte eria ssifie oteob teob Bact d; ia;Pr hapro ; U acter ;Alp TM6 ncla B cteria ria; ssif eoba acte ied; ;Prot B ; U cteria ified ncla Ba lass ssifi Unc ed; d; B ifie acte lass ria; Median: ~6700bp nc The U ; U rm fied nc oto ssi lass gae cla ifie ;The Un ; B d; rm ied ac otog ssif teri ae classified in PacBio data and absent cla a;T ;Th n s; A er U bu Ba 06; mot iglo cte oga err ria les );T ;Ca ;Th 1 us; B nd erm oup cill act ida oto gr ba eri te gac ub eni a;C div eae (S a s; B an isio ;Fe Figure 1 (Right): A diagram of the eae e;P illu ac did n O rvi a c t a d riac ce ba eri te P8 oba te illa ni a;C div ; cte bac ac ae s; U an isi rium Date: Current ido nib e;P illu nc did on ; c e a c la a O s;A ;Pa ce ba ss te P8 ale les illa mo ; ifie div ; eri illa ac er us Ba d; isi act ac nib ;Th cill cte on ob li;B ae ae iba ria OP cid cil s;P ce n ; U ;C 8; ;A Ba le illa ae us n an eria es; illa ac ;P ill cla did ct ut ac ib ae ac ss at from environmental samples show high concordance ba ic ;B n ce ;B if e ido irm illi ae lla ae s; B ied di Ac ;F ac s;P ci e llu ac ; vis ia; ria s;B le iba lac ci te ion ter cte te illa en cil a ria O ac Ba icu ac a a e;B ; B ;P P ob rm i;B s;P s;B a us ac lan 8; cid ;Fi cill le le ce cc te ct from iTags. Branch colors denote phyla ia;A ria Ba illa illa illa co ria om ter cte s; ac ac ac to ; B ;P yc ac Ba ute li;B li;B ;B ep lus ac lan et B ic cil cil les tr cil te ct es irm a a lla e;S a ; ria om ;P ;F s;B s;B ci a eb us B ;P y hy ria te te Ba ce m in ac la ce cis cte icu cu lli; ca Tu os te nc te a rm mi ci oc e; or ; ria to s;P a B Fi ir Ba c a p us B ;P my h er ia; a;F s; to ce os in ac la c yc ae generation of a Read-of-Insert er ri te rep lla lf s te n et is ;P t te u t i u o r c e p h ac c ic ;S ac es or s; B ia to s; ha yc B Ba rm s b D p u a ;P m Ph e is Fi le lo e; os in c la yc y ra ph ia; illa yc a lf s te n e cis e;P ae er ac lic ce su ro ; ria ct te p h ra ct b ;A a e po m U ;P om s; ha yc le a to s cc ;D s lu nc l y Ph er is s; B ac ale o e lfo u la an c y ae ph Ph i;L ill oc ea u ac s c et ci ;M a yc ill ac pt c s m ; B sif to es sp S er is ac B e ca e o ed ac ie m ;P h B ale ph B lli; ;P c ;D ot ifi te d yc h ae L s a s; ci s co ae lf s r ; e yc ra 9; ;P era te a le o su s ; B ia te is e hy c cu ;B ia pt ce e la d ac ;P s p ;M c ea i es rid e ca D nc fie t la ;P ha S isp e irm ut st ;P c e; U si er n h e B h ;C ;F ic lo s co a s a; U ia ct yc ra L9 ae L5 ia m ;C le to e la n ;P o is e ; ra 00 er ir ia ia p ac c sm cl l m p ;M ce − ct ;F d rid e c n a a an y ha S a 3; a ria tri t P c U pl ; B ss c ce e B e; B te s os s; co e d a if to te ra L C c lo l le o ol ie c ie m s e 9; L5 . a ;C ;C ia pt h if te d y ;P ;M 00 B s ia d e c s B r ; c te id tri P A s ; a ia e la S −3 u tr s s; ; la ed c ;P te nc BL ; ic s lo e e c fi te l s t 9 m lo ;C al ea n si B r an ;P om ; ir ;C a di c U s 1; a ia c l ;F s di ri ta la 4 c ;P to an yc ia te ri st a c B t m c e er u st o n B er la to ta ct ic lo l sm U P ; a ia n yc m c a rm C ;C a O d c ; ct e y ia B i ; ia pl a; fie t P o t c ;P consensus from an unrolled ;F es id e ri i U e la m es e l a t r l s ; r t a ri st o te s 1 n ia nc yc ;P a n te icu lo h la 4 c ;P t e la ci ct c m c ac c B la l o t n a; o a ir ;C ;A b n P ; Uncla s a m es c P m B ;F s s o U d s nc y ; to la y a te le in ;O ie if c P m n c ri u a ct a if ie to e la c et e ic at A ri s ; U s d m te n yc to al act rm m ; te s d n s ; y s c e m es B i s ia c la ie c ified c ;P to ta y ;P ;F la er a c if ; U la e la m c ce l ria p t b n s 1 n s te n y ia t an e le c o U s 4 c s ; s c c ; al c • SMRT® Sequencing could be a robust, cost-effective t o a n la l if ; t e P e t ac h B ti c B ; U a ie P o ta la s; om B c c n P d n s d la m c n P A A O e c s ; n y ia c la yc s; ; U ; fi l if c c ; to n e e ia ia i ; U a ie to e P m c t t r r s 1 n s d m ta la to ac u te e s S c s ; c n yc m e ic c t a l if y ia c e a ll a l J ; U a ie c ; t t yc e o ac c n 1 n s e P o a e ;u M B b n o S c s d ta l m le t n ; o U i l i ; c a y s a c  U Reverse #1 J f n s n is ; a ie i c c ;P ce u te ti v n 1 n s a t e a lt u i S c s d ;P o ta la e u ic c d io U l i ; l m l n ; re r ;A s J ; a fi an e c u d e e i n 1 n s e y s to n ; n ia t iv S c s d c ce ;P m cu e r a io J ; U l i ; t l T e d d s a fi o ta a y ltu ; t i e i 1 n s e m l n c r ia c d t v n S c s d y e c e e r a n a i io ; U l i c s t ta d te d J I a f ; e ;P o c ; c B a id is y n s ie t m e a C d te v n il c s d al la y a ; i io ; U l n c sequence consisting of 1 full and 2 B a n a d m n a if ; e c e e i s a i s ; r a id i a t s e t t u e e v e U cl s d ;P o a n t ;C d t i F ; a i l m c c c a n a d I; a d n fi ; a y e u a i a d h e c ss e n c a lt r i e n c i ; Un c e e u B e C d t o if la i d t ; r t ; a i o d f ; o ta u e c a n t ir s e B c s ie m c n d a i a id c s fi ; l s d e c ; r e p a i d a a i y a u B e ;C d s S l s c s fi ; c e lt t n ; c ie ; B e e ; u c ia a b s f a te s d tac u r a r u e n a i d i n e C a l s e U c r fi ; c d B e ; S U c s i ; t i e e ; t ; e f d n e a a u c ia a c n a i ; d e lt a r i l s e ; U c r C ; u r ta U c i l i ;un r B te e s if d n a a a e e n a ; U c ; d c a l s ie s C n c ; a ct U c s f 3 n l s d u a h i B a a l B n a L ; c s if i tu b c l s i n d o U c s B 3 a la s e r o r U c d a e n i n a S L ; s if d i t d l d n t i d e p c B B e s e ; ; U M ; c a ya S n ; S ie r i d d ; f d a l i f t C i B a a ie ; e i ; s U ia e ; c v M s a s ; a e r ; r d B t d i i l s ; e s D d s r e a u e c ; a i t i a r i e iv i e t a r a l f d B t f o t l i ; c e ia i i e h e c u e a e f s n c s i 4 ; B t r e a c f e i ; d i a p a n s i 6 c a D r o O h e ; B a r b s h n s t r ; n c i U a A a c e i ; e i P t u l s B a C b o p m ; a o ; e t r i O 3 n r n s c a H c e ; n a i u U a i h method for multiplexed strain identification i e l c C ; e n l m ; t a o c p t n i r P a c i B n c e l ya L a u d ; i ; h o c t partial pass “subreads” n U l t a B e 3 S ; e n h B d c i a c e r r C ; e e d ; i l o ; ; e c p i B ; a o o r s a c h f a c l a L U i e d ; a r B c e a a ; a v i i U i c r b Table 1 (Below): Counts of clusters and i e o p ; ac c t ; c t r l ; d f e d a a o s r l i i s i B e s B e c i ; t i B n t ; u ; e e e i f e d s ; c e b e c d s s ; e r C D a a i i B a B a a i d c t i s a r e f e ; B a i t r i v a a a s er r a ct s i i d U f ; ; c c e o e i t l c l e a l a h − h f B h e t c d ; a i i C i B c a a s i a i s e ; B t a ; e t c Ba r i f c u y d Unclassified; Bacte a l f ; l T i n c c p c ; e i B e d o e a ; l d t i e a t s o B i n s s f c a; d h o a d a ; i n c i e h a t s l i e t B h r c r e r y f d r r h s i e s i t s d a i c e e l i u a m t ; i c r p i e o d t V f r i o o e n c s c l cte l e p h U i B t ; s i e a i i e f t a c p e r a a s u r i f t i f t i e e b b n P i l r r r s t e a s n c i r o s O s i i i d ; U a e ; t f act i e c o m S s ; s r a a e c l s i B e t s i e M i e t u ; n a r ; e s s s a i e a r e l U l B t d b t r s ; c a c a s i ; e ria d e L n s ria;B r i e n l u i l s a a S ; i B ; e s M a t ; t s U c a i t a i o n l a e C l s a e l a B a ; e r e a s e f c u t b l a ; c ; a l u U c p n ; i a i c l o i t a ; B s C

s a ;B ; L t e c r ; h B d r c e e r o U n c a t a B ; a l h C D e c c r c n e oi i ; a u U t a a d e l e n c d y n s v c B a i a t b t e o U a e s i c c n h t ac n h i acter r t t i d U c ; t e de e t ; u r r i a n c r o a ; c U e e S o i o c e c ; g l o n D c U t o o e r t a i d d t i c c e p o m n t s r e s t i U ter o d B o i e r b o r e i t e n o t i y ; h B V o r d e c e s o d i e e b i c ; e r v − r f l r d p c a o i a a a o e s ; a b e c n o d r o

oid t a a oidet 1 o v i a l S i u i e e o g r i t i ; c i e B C a d e c ; i d a r r d e v a ; r n d ; t s i r C i t s c i a b e e t e i d a l b ; ad i u s C e s d e e e ; n e r e i ; et ; i a o t s B i i ; r S ; a e t ; e a r s T e P t s H C d s S n i V e ;  e Reverse #2 t h ; i i c es

es;Sp D i J ; a S ; ; e c s ; n H a h c s y i C S p A B t p e d s A e e a ; ; e a p 2 H t s i ; S S a ; h A 1 a S h e ;Sp o i S p r r c − m r t S B t h − i A a i 7 c p l m p 1 p 2 e e S a h n o u p n i 2 t i p l d t t n 1 7 ; h h e i 8 h i i a r g e r a e c h n ; e h t hi o hing g ; a i s c ; i 7 L p r e o r n l a a i n g e t i b ; ; n o ; o g t n S A e ng b g g o s B O c i r ; g i e b a ; c g d a a o ; A o b a e a a T o l a c a n o a b i oba e r i obac c R b I c i a b b e h c b l s a e b n t 4 a c e reads at various levels of assignment for e b e a e y I a t ; c a a o l e s 0 c t C i n o r e r r c e t h c r m cte r i ; e 6 y t y i ; c l t P y e r i p t a C i m i r c

teriia; e i r i e i c a s a i c a r i ; o i o r a a l h c r l m m i S t rii ; i o a i r m F i S l l i e a ; i n o ; a C i p e o a p S b d ; a a;S a ; e c s F d p h s r ; h a e i s i ; S ; p o e S ; L u S h y ( Sphi c c i C ; r d s e p h n r b ; M i l r p i t M p a e ph n o A i h i i h g e d a h n u e ; a r i h g m i e S i o r e r t m i r a V n g i e i

in n o i i a i i t ; n b o n ngobac o e o n e S r g ; c t g b c a g a o go n t e i a o b e h c a t e o a e r t o c a p r b a e g a a b c B e a c t h b d t e t ba c r r a e e b t a r a a m a a o e c c t c r ; e e c g e b c u c i u r a n cte t a u I c n t r a p e i e o t n a terial e s B i e l y r n a e a c n r c ; A I r l r i r i e ria e e l e t i a s u i a e r o ) a s a d c ; ; l ; l t l P r e e s W l ; e C

les e u e A W ; ; d es;W s a ; ; W r O r s s C ; a ; l e i a ; B C u i ;

s ;W O r W H d i W C ; r d 0 H d e ; e B s t H i C 1 i t C b C e CHB B c 1 d c R B a H H 1 S a HB − e a c 1 0 Figure 2 (Top-Left): Raw read B − b e S B 6 t B − 1 e

o 6 1 9 a 1 1

1 6 2 r t e 9 Illumina iTag and PacBio 16S data. Data n − ; ; r i − − − ; 9 a t ; t 6 e 6

6 ; 69 r c c 9 9 e 9; A n ; ; ; I c ;

n a s I i s r a s e l t s

c C a References l ; a

a b i C ;

o b a

i n o i

r b t c i o c r c A m i ;

o a

i c m r u o r e c t r

u c e r r a V ; e B Reverse #3  a i V r ; a e i t r c length distributions for sequencing e a t c B sets show strong concordance with each a B of full-length 16S (~1500bp) other except at the OTU level, where the [1] Rinke, Christian, et al. "Insights into the phylogeny and coding potential of microbial dark amplicons. The dotted vertical line Reverse #4  matter." Nature (2013). denote the minimum read length for Grouping iTags PacBio 16S [2] https://bitbucket.org/berkeleylab/jgi_itagger the generation of Reads-of-Insert Reverse #5  consensus. # of phyla resolved (reads) 32 (4,407) 34 (5,000) [3] https://github.com/PacificBiosciences/rDnaTools # of candidate phyla (reads) 18 (2,139) 16 (1,724) Reverse #6  [4] Letunic, I., & Bork, P. (2011). Interactive Tree Of Life v2: online annotation and display of Figure 3 (Bottom-Left): Mean phylogenetic trees made easy. Nucleic acids research, 39(suppl 2), W475-W478. quality-values for full-length 16S # of families resolved (reads) 45 (3,906) 49 (4,636) Reverse #7  CCS sequences by raw (unrolled) [5] Cole, J. R., Q. Wang, J. A. Fish, B. Chai, D. M. McGarrell, Y. Sun, C. T. Brown, A. Porras- sequencing read length. Box-width # of families from candidate phyla (reads) 22 (3,197) 18 (3,217) Reverse #8  Alfaro, C. R. Kuske, and J. M. Tiedje. 2014. Ribosomal Database Project: data and tools for denotes relative abundance. # of OTUs (reads) 1,843 (4,407) 359 (5,000) high throughput rRNA analysisNucl. Acids Res. 41(Database issue):D633-D642; doi: 10.1093/nar/gkt1244 [PMID: 24288368]

Pacific Biosciences, the Pacific Biosciences logo, PacBio, SMRT, SMRTbell and Iso-Seq are trademarks of Pacific Biosciences of , Inc. All other trademarks are the property of their respective owners. © 2014 Pacific Biosciences of California, Inc. All rights reserved.

Top View