LVTree Viewer: An Interactive Display for the All-Species Living Tree Incorporating Automatic Comparison with Prokaryotic Systematics

Guanghong Zuo，Xiaoyang Zhi，Zhao Xu，Bailin Hao*

?

LVTree Viewer: An Interactive Display for the All-Species Living Tree Incorporating Automatic Comparison with Prokaryotic Systematics

Guanghong Zuo1，a，Xiaoyang Zhi2，b，Zhao Xu3，c，Bailin Hao1，*，d

1T-Life Research Center，Department of Physics，F(xiàn)udan University，Shanghai 200433，China
2Yunnan Institute of Microbiology，Yunnan University，Kunming 650091，China
3Thermo Fisher Scientific，South San Francisco，CA 94080，USA

Received 28 October 2015；revised 4 December 2015；accepted 21 December 2015 Available online 25 March 2016

Handled by Jingfa Xiao

KEYWORDS

Archaea and Bacteria；

16S rRNA phylogeny；

Monophyly；

Collapsing and expanding a tree；

Lineage modifications

Abstract We describe an interactive viewer for the All-Species Living Tree（LVTree）. The viewer incorporates treeing and lineage information from the ARB-SILVA website. It allows collapsing the tree branches at different taxonomic ranks and expanding the collapsed branches as well，keeping the overall topology of the tree unchanged. It also enables the user to observe the consequence of trial lineage modifications by re-collapsing the tree. The system reports taxon statistics at all ranks automatically after each collapsing and re-collapsing. These features greatly facilitate the comparison of the 16S rRNA sequence phylogeny with prokaryotic taxonomy in a taxon by taxon manner. In view of the fact that the present prokaryotic systematics is largely based on 16S rRNA sequence analysis，the current viewer may help reveal discrepancies between phylogeny and taxonomy. As an application，we show that in the latest release of LVTree，based on 11，939 rRNA sequences，as few as 24 lineage modifications are enough to bring all but two phyla（Proteobacteria and Firmicutes）to monophyletic clusters.

E-mail: hao@mail.itp.ac.cn（Hao B）.

aORCID: 0000-0002-7822-5969.

bORCID: 0000-0003-2897-888X.

cORCID: 0000-0002-2978-7025.

dORCID: 0000-0003-3547-564X.

Peer review under responsibility of Beijing Institute of Genomics，Chinese Academy of Sciences and Genetics Society of China.

http://dx.doi.org/10.1016/j.gpb.2015.12.002

1672-0229?2016 The Authors. Production and hosting by Elsevier B.V. on behalf of Beijing Institute of Genomics，Chinese Academy of Sciences and Genetics Society of China.

This is an open access article under the CC BY license（http://creativecommons.org/licenses/by/4.0/）.

Introduction

The All-Species Living Tree project（www.arb-silva.de/ projects/living-tree）isaninitiativeofthe Editorial Officeof Systematic and Applied Microbiology in collaboration with the ARB-SILVA（www.arb-silva.de）and List of Prokaryotic Names with Standing in Nomenclature（LPSN，www.bacterio. net）projects to construct a single phylogenetic tree for all available type strains of Archaea and Bacteria based on high-quality 16S rRNA sequences［1-4］. The latest releaseLTPs123（as of September 2015）contains 449 Archaea and 11，490 Bacteria sequences. However，it is a hard job to comprehend such a big tree by scrolling up and down the long PDF，especially when one wishes to compare the phylogeny with taxonomy at all ranks from phylum down to species by checking the monophyly of all the individual branches.

A similar situation has been encountered while working on the whole-genome-based Composition Vector Tree（CVTree）web server［5-8］in order to enable the server to deal with many thousands of prokaryotic genomes in one run. To address this issue，several distinctive features have been introduced in the latest CVTree3 web server［8］. These features include（1）collapsing a monophyletic tree branch into a single leaf when the branch represents sequences from the same taxon according to a reference taxonomy；（2）expanding a collapsed leaf to show the constituent leaves，while the overall topology of the tree is preserved during collapsing and expanding；（3）reporting taxon statistics，i.e.，the number of sequences represented by each taxon，monophyletic or non-monophyletic，in the form of a hierarchical list；（4）enquiring a designated taxon by entering its name in a“Search Query”box（this is the quickest way to get to the point of interest；while the neighborhood of the enquiry is properly expanded，all the other branches are maximally collapsed）；（5）making trial lineage modifications and displaying the result of re-collapsing with renewed taxon statistics；and（6）outputting print-quality sub-tree figures in different formats.

All these features have been transplanted to the LVTree Viewer presented in this study. There are some technical points in the implementation of the Viewer. For more details，the users are advised to consult the CVTree3 User’s Manual at http://tlife.fudan.edu.cn/cvtree3/.

Description and usage of the LVTree Viewer

The LVTree Viewer is freely accessible at http://tlife.fudan. edu.cn/lvtree/ without login requirement. It takes the tree structure from LTPs123_SSU.tree.newick and lineage information from LPTs123_SSU.csv，both downloadable from the All-Species Living Tree project website. In fact，two previous releases（LTPs121 and LTPs119）are also accessible on a rotating basis.

By clicking on“Show LVTree”，a maximally-collapsed tree with two leaves，＜D＞Archaea｛422+7｝and＜D＞Bacteria ｛11156+334｝，appears in the next screen. The taxon｛n+m｝notation is explained appropriately in what follows. Lineage information is available for each 16S rRNA sequence in LVTree. When hovering the mouse over a taxon name，a small popup window with lineage information appears for a few seconds. For example，an entry may read as“＜D＞Bacteria ＜P＞Actinobacteria＜C＞Actinobacteria＜O＞Streptomycetales ＜F＞Streptomycetaceae＜G＞Streptomyces＜S＞Streptomy ces_longisporoflavus＜T＞Streptomyces_longisporoflavus__D Q442520...”，where＜D＞，＜P＞，＜C＞，＜O＞，＜F＞，＜G＞，＜S＞，and＜T＞stand for Domain，Phylum，Class，Order，F(xiàn)amily，Genus，Species，and sTrain，respectively. The line above summarizes the correct lineage of a well-studied streptomycete species Streptomyces longisporoflavus，which was initially described by Waksman and Henrici in 1953. It will also be shown as an example of lineage modification in the later part of this study.

A standard specifier“Unclassified”is used to indicate any rank in classification lacking taxonomic assignment，e.g.，＜O＞Unclassified＜F＞Unclassified＜G＞Vampirovibrio gives a genus name without order and family assignment. Lineage information containing at least one specifier“Unclassified”is considered incomplete. If there are“n”entries with complete lineage information and“m”entries with incomplete or missing information for a given taxon，these numbers appear as taxon｛n+m｝. Furthermore，the two lines＜D＞Archaea｛422 +7｝and＜D＞Bacteria｛11156+334｝are shown in red，because both are monophyletic，whereas non-monophyletic entries would appear in blue. An entry can be expanded by clicking on the solid circle in front，whereas an entry with no circle present contains only one single sequence and thus cannot be further expanded. In a screen with a partially-expanded tree one may click on a blank circle to have the subordinate branches collapsed.

In the preamble line of a display screen，four boxes are labeled as“Top Node”，“Lineage Modification”，“Monophyly List”，and“Search Query”. By clicking on“Monophyly List”，a hierarchical summary of all taxa with statistics appears with“Yes”denoting a monophyletic taxon and“No”denoting a non-monophyletic taxon，respectively. There are four options to present the summary:“Total”shows statistics of all taxa；“Monophyly”gives a list of all monophyletic taxa（since any taxon represented by a single sequence is trivially“monophyletic”，one can hide such entries to get a more comprehensible list）；“None”lists the non-monophyletic taxa；and“Unclassified”shows entries with at least an“Unclassified”specifier in the lineage information，i.e.，those entries that correspond to the augend“m”in the taxon｛n+m｝notation. The quickest way of getting to a taxon of interest is using the“Search Query”function. For example，typing“Epsilonprote obacteria”to replace“Search Query”in the box would lead to a collapsed tree with＜C＞Epsilonproteobacteria｛111｝displayed in green，indicating that this class represented by 111 rRNA sequences has already been well defined，as shown in Figure 1. The“Top Node”pull down list is used to restore the displayed tree to a designated node as the leftmost node after collapsing.

The present prokaryotic taxonomy as collected in Bergey’s Manual［9］and LPSN［10］is largely based on 16S rRNA sequence analysis. Ideally speaking，one should expect full agreement of the branching order in LVTree with the taxonomic hierarchy. Given an input dataset and a well-tested method of phylogeny inference，the LVTree must be considered as a fixed subject that cannot be adjusted or modified. To the contrary，prokaryotic taxonomy has always been a work in progress. Therefore，discrepancies revealed when comparing LVTree with systematics most probably hint on problems in taxonomic assignments. However，there might be errors such as mislabeling of sequences in the underlying LVTree dataset. The LVTree Viewer may help spot some of the errors as well.

The LVTree Viewer provides a mechanism to examine the consequence of trial lineage modifications. A user may submit a Lineage Modification file，in which each entry describes a proposed modification: old_lineage＜space＞new_lineage.

When suggesting a lineage modification，the names under ＜S＞and＜T＞are always kept unchanged. In doing so，no confusion would occur concerning the published or cited species names，which can always be used to do“QuerySearch”. In other words，species and strain names in a collapsed tree appear in the same way as in the original LVTree.

By clicking on“Lineage Modification”an empty window appears. A user may type in own modifications or drag in a prepared file from one’s local computer. Alternatively，an example Lineage Modification file（File S1）may be invoked. Upon clicking on“Submit”，it takes a while for the system to display the re-collapsed tree and the re-calculated Monophyly List. This process may be carried out repeatedly and one can“Save”the Lineage Modification file to the local computer for future use. By clicking on“Clear Text”and then“Submit”，one may undo the modifications. The Project number given at the beginning of the job session becomes useful，when the user has formed a Lineage Modification file. By typing in the Project number when starting a new LVTree session，the Lineage Modification file is called back from the user’s Workspace，which is kept for 7 days after the last run.

When accessed for the first time，the LVTree Viewer comes with a default Lineage Modification file（File S1）. In order to check the modifications suggested in the subsequent Sections，the user should disable this file by using the“Clear Text”and“Submit”buttons. In other words，the following discussions aim at a“bare”viewer without making any lineage modification.

Figure 1　A collapsed All-Species Living Tree with the neighborhood of Epsilonproteobacteria expanded

Lineage modifications that bring all but two phyla monophyletic

The notion of monophyly serves as a basic guiding principle when comparing phylogenetic tree with reference taxonomy. As an application of the LVTree Viewer，we check the monopoly of all branches at the rank phylum.

An inspection of the Monophyly List shows that 21 phyla appear as monophyletic clusters without making any lineage modifications. These include Aquificae｛27+1｝，Armatimonadates｛3｝，Caldiserica｛1｝，Chlamydiae｛13｝，Chloroflexi｛33｝，Chrysiogenetes｛4｝，Crenarchaeota｛51｝，Cyanobacteria｛15｝，Deinococcus-Thermus｛84｝，Dictyoglomi｛2｝，Euryarchaeota ｛371+6｝，F(xiàn)ibrobacteres｛4｝，Gemmatimonadates｛1｝，Lentisphaera｛4｝，Nitrospira｛7｝，Planctomycetes｛23｝，Spirochaetes｛92+1｝，Thaumarchaeota｛1｝，Thermodesulfobacteria ｛8｝，Thermotogae｛43｝，and Verrucomicrobia｛41+2｝.

By typing the non-monophyletic phylum names into the“Search Query”box one by one，it is easy to spot a small number of outliers that violate monophyly of the corresponding phyla. Examining the neighborhood of a given outlier，in combination with tracing back the history of taxonomic assignment，helps to spot a sequence error or suggest a proper lineage modification.

To demonstrate this function，“＜P＞Actinobacteria”is typed into the“Search Query”box. A big cluster in green with a single branch in red appears when scrolling down the tree display. This represents the main body of the phylum Actinobacteria with an“invading”species“＜S＞Flavobacter ium oceanosedimentum”from the phylum Bacteroidetes（data not shown）. At the same time five species escape from Actinobacteria into other phyla，including＜S＞Streptomyces longisporoflavus as shown in Figure 2.

Apparently，a problem is encountered. The genus Streptomyces and the species S. longisporoflavus were proposed in 1943 and 1953 by Waksman and coworkers. Nowadays Streptomyces is a large genus containing 669 species and 38 subspecies as listed in LPSN［10］. Being a long-studied species of medical significance，the taxonomic assignment of S. longisporoflavus is beyond all doubt. Morphologically S. longisporoflavus is clearly distinct from Brevundimonas species，so the confusion can only stem from the 16S rRNA sequence （GenBank accession No. DQ442520）deposited by Goodfellow and colleagues in 2006. Fortunately，they deposited another 16S rRNA sequence（GenBank accession No. NR_115963）for this species in 2015. A BLAST comparison shows that NR_115963 has high similarity with 16S rRNA sequences from other Streptomyces species，while DQ442520 is close to 16S rRNA sequences from Brevundimonas. Most probably，DQ442520 represents a mislabeled Brevundimonas. SinceLVTree users cannot change or replace the sequence database，we propose the following lineage modification just to“hide”the inconsistency:

＜P＞Actinobacteria＜C＞Actinobacteria＜O＞Streptomy

cetales＜F＞Streptomycetaceae＜G＞Streptomyces＜S＞

Streptomyces_longisporoflavus

＜P＞Proteobacteria＜C＞Alphaproteobacteria＜O＞Rhi

zobiales＜F＞Claulobacteraceae＜G＞Brevundimonas

＜S＞Streptomyces_longisporoflavus（please note that in

between the two seemingly lines above，there is only

a＜space＞character，not a carriage return）.

Figure 2　A sequence with erroneous lineage information appeared under the name of Streptomyces longisporoflavus

Similarly，Thermoleophilum minutum，another outlier from Actinobacteria，also has two different 16S rRNA sequences present（GenBank accession Nos. HQ223108 in LVTree and AJ458464 in Vol. 5，the Bergey’s Manual［9］，respectively），leading to inconsistent taxonomic placement. This is treated analogously by a temporal lineage modification（File S1）. The genus Thermoleophilum was proposed in 1986 with two species described so far（Thermoleophilum album and T. minutum）. New family，order，and class were introduced for this genus in 2005，2009，and 2013，respectively. In LVTree，while T. album remains in the phylum Actinobacteria，T. minutum gets deeply into＜G＞Pseudomonas in＜P＞Proteobacteria.

Another apparent sequence error happens to Filifactor alocis，a Firmicutes species getting into the very depth of the phylum Fusobacteria（Figure 3）. The 16S rRNA sequence （GenBank accession No. X55406）chosen by LVTree was published among a group of 14 members of the genus Fusobacterium in 1991 by Lawson et al［11］. However，it was soon pointed out that“Lawson et al have been erroneously sequenced some other Fusobacterium species instead of Fusobacterium alocis 35896T”［12］and the species was reclassified as Filifactor alocis to the phylum Firmicutes. In other words，X55406 is associated with some other strain of Fusobacterium，but Lawson et al simply changed its name to Filifactor alocis when submitting the 16S rRNA sequence in 2004，causing the inconsistency revealed here. Therefore，the LVTree dataset should exclude X55406 from the Fusobacterium group and use the sequence with GenBank accession No. AJ006962 （indicated in Vol. 3，the Bergey’s Manual［9］）for Filifactor alocis. By the way，AJ006962 still carries a wrong sequence name Fusobacterium alocis but with a correct organism source Filifactor alocis. The suggested lineage modification is a compromise before appropriate correction is made in LVTree.

We skip further discussions and list briefly all suggested lineage modifications which are enough to bring all but two phyla （Proteobacteria and Firmicutes）to monophyletic clusters.

1. Arthrobacter is a genus proposed in 1947 and the species Arthrobacter viscosus was first described in 1965. It inherits the phylum assignment to Actinobacteria from the genus name，but in LVTree it gets into the depth of the genus Rhizobium within the class Alphaproteobacteria. Its taxonomic position was questioned by Keddie et al in 1986（Vol. 2，the 1st Edition of Bergey’s Manual）. Its 16S rRNA sequence that was published in 2005［13］showed a high degree of similarity to several Rhizobium species（97.6%-98.7%）. The suggested lineage modification takes these facts into account.

2. Brevibacterium is a genus proposed in 1953 and the two species Brevibacterium halotolerans and Brevibacterium frigoritolerans were described in 1967，inheriting the phylum assignment from the genus name. In LVTree，these two species locate unquestionably in the genus Bacillus，far apart from the main cluster made of 30 Brevibacterium species. In fact，when initially submitted in 2007，16S rRNA sequences（accession Nos. AM747812 and AM747813）were assigned a Bacillus lineage in GenBank. Therefore，the suggested lineage modifications just recollect this overlooked fact in the LVTree database. As a result，these modifications lead to a monophyletic＜G＞Brevibacterium｛30｝in LVTree.

3. The genus Acetobacterium and species Acetobacterium flavidum were proposed in 1984［14］. A second species Acetobacterium faecale was proposed in 1987［15］. However，it was stated on page 43 of Vol. 4，the Bergey’sManual，that“placement of Acetomicrobium as a member of the family Bacteroidaceae is not certain because of the lack of rRNA gene sequence data.”In fact，the type strain of A.flavidum bears 99.7%（［4］）or 99.8%（［16］）16S rRNA sequence similarity with Anaerobaculum mobile，the sixth sequenced species in a newlyproposed phylum Synergistetes［17］. In addition，the 16S rRNA sequence of A. faecale“shows an unexpected affiliation with a deep branch in Clostridia（99.46%similarity against Caldicoprobacter oshimai）”［18］. The suggested lineage modifications take these facts into account and effectively render Acetomicrobium an empty genus.

Figure 3　Clostridium rectum and Filifactor alocis violate monophyly of the phylum Fusobacteria

4. Anaerorhabdus furcosa was reclassified from the genus Bacteroides in 1986［19］as the only member of a new genus，but the phylum assignment of the latter has not been questioned then. Therefore，it is still classified under Bacteroidetes in the original LVTree lineage. However，previous study indicates an unexpected affiliation of this species with the family Erysipelotrichaceae （91.6%similarity with Holdemania filiformis）［4］. In fact，in LTPs123，two other neighbors of A. furcosa，Solobacterium and Bulleidia，appear even closer to Erysipelotrichaceae. The proposed lineage modification reflects this fact.

5. Bacteroides cellulosolvens has been reclassified to a new genus in the class Clostridia as Pseudobacteroides cellulosolvens in 2014［20］. However，both B. cellulosolvens and P. cellulosolvens appear juxtaposed to each other in LVTree. It seems that it is appropriate to delete the one misclassified to Bacteroidaceae.

6. Bacteroides coagulans，a species described in 1933，locates deeply within class Clostridia of phylum Firmicutes，far from the phylum Bacteroidetes in LVTree. In EzTaxon-e［21］it is assigned to a newly-proposed genus Ezakiella［22］under Incertae sedis family XI. Our suggested lineage modification follows EzTzxon-e with ＜F＞Unclassified，though the re-collapsed LVTree hints on its belonging to the family Peptoniphilaceae.

7. The description of both Bacteroides galacturonicus and Bacteroides pectinophilus in 1986 was based on phenotypic data［23］. The 16S rRNA sequences were deposited to GenBank（accession Nos. DQ497993 and DQ497994）in 2006 with mentioning of“reclassification”in the unpublished title，but only one lineage has been changed to Clostridiales. Their positions in LVTree show that these strains should be assigned to the family Lachnospiraceae in two different genera. This fact is reflected in the suggested lineage modifications. It is of note that the whole-genome sequence（GenBank accession No. ABVQ01000036）instead of the 16S rRNA sequence（DQ497993）was given in the strain tag of B. pectinophilus in LVTree.

8. Type strain of Flavobacterium oceanosedimentum shows an unexpected 99.8%similarity to Curtobacterium citreum［4］. Although the association of this species with the genus Flavobacterium was questioned in the 1st edition of the Bergey’s Manual in 1984，formal proposal to transfer it to the genus Curtobacterium in the phylum Actinobacteria appeared only in 2009［24］. Such an assignment was adopted in Vol. 5，the Bergey’s Manual ［9］，which is reflected in the proposed lineage modification. The modification leads to a monophyletic branch Curtobacterium｛8｝in LVTree，and the monophyly would be violated if F. oceanosedimentum was missclassified to Flavobacterium.

9. The 16S rRNA sequences of the two Caldithrix species were deposited in 2003［25］（Caldithrix abyssi）and 2010［26］（Caldithrix palaeochoryensis），respectively，without proper lineage information except for the genus name. Their strain tags in LVTree carry a specifier“U nclassified-Deferribacterales”，but actually they take the position of a separate phylum in LVTree（Figure 4）. On the Integrated Microbial Genomes（IMG）website （img.jgi.doe.gov），they are given a new phylum name Caldithrixae，not listed anywhere else in the literature. The suggested lineage modification adopts this phylum assignment with“Unclassified”lower ranks.

10. In LVTree，the 16S rRNA sequence（GenBank accession No. RF733681）of Ilyobacter delafieldii escapes from the phylum Fusobacteria and appears as a close sister to Clostridium homopropionicum with 98.8%sequence similarity as described previously［4］. The latter is attributed to the genus Clostridium senso stricto. Therefore，thesuggested modification puts the species I. delafieldii in the Clostridium lineage.

11. The species Fusobacterium naviforme was described in 1909 and the genus was introduced in 1922. In LVTree F. naviforme appears to be the closest neighbor of Moryella，a single-species genus in Lachnospiraceae in the phylum Firmicutes. The lineage modification follows EzTaxon-e［21］and assigns F. naviforme to the genus Moryella.

12. Clostridium rectum was attributed in 1922 to the old genus Clostridium Prazmowski 1880，long before the rRNA concept appeared. In 1994 it was recognized that“although C. rectum forms endospores，it is clearly a member of the genus Fusobacterium”［27］. The 16S rRNA sequence was deposited in 2004 with a conditional organism name［Clostridium］rectum. The suggested lineage modification simply recognizes this consensus. After making the modifications explained in this and the previous entry，there appears a monophyletic genus Fusobacterium｛20｝in the re-collapsed LVTree.

13. The genus Acetobacter was described in 1898 with type species Acetobacter acetis discovered by Pasteur in 1864. The bacterium Acetobacter pasteurianus subsp. ascendens has been well-studied since 1898. In LVTree it locates within the genus Lysinibacillus in the phylum Firmicutes，while the other two subspecies strains of Acetobacter pasteurianus remain in the designated phylum Proteobacteria. The suggested lineage modification assigns it to Firmicutes.

14. Gemmiger formicilis is a species described in 1975［28］. The 16S rRNA sequence of the type strain bears 97.5%similarity with that of Subdoligranulum variabile in the family Ruminococcaceae［4］. Both Gemmiger and Subdoligranulum are single-species genera for the time being. The suggested lineage modification temporarily puts G. formicilis in the genus Subdoligranulum.

15. The bacterium Vampirovibrio chlorellavorus was first described as a member of the genus Bdellovibrio in 1972［29］and a new genus was proposed for it in 1980 ［30］. In LVTree，it is next to the phylum Cyanobacteria｛10｝，as if it was a cyanobacterium or represents an yet uncharacterized phylum. Although the suggested lineage modification attributes it to an“Unclassified”phylum，LVTree Viewer’s re-collapsing mechanism automatically absorbs it into Cyanobacteria. In view of the insufficient representation（only 15 among the 11，939 sequences）and the even less-studied taxonomy of Cyanobacteria，this problem awaits further scrutiny. In addition，we note that the order Bdellovibrionales turns out to be an outlier fromthe phylum Proteobacteria in whole-genome-based CVTree［8］.

Figure 4　A single LVTree branch accommodating 8959 16S rRNA sequences

16. In LVTree，both the single-species genera Asteroleplasma anaerobium［31］and Haloplasma contractile［32］fall in a“grey area”between the Firmicutes and Mollicutes. Mollicutes was taken out of Firmicutes as a separate phylum in Vol. 4，the Bergey’s Manual［9］. This area predominantly consists of species from Erysipelothrichaceae，the only family in the class Erysipelotrichia. As for A. anaerobium，“the question of possible monophyly and the phylogenetic position of Asteroleplasma with respect to other Mollicutes remains open”（Vol. 4，the Bergey’s Manual，page 723）. The branching position of H. contractile“was equidistant to the taxa considered to be representative lineages of the phyla Firmicutes and Tenericutes”［32］.

17. In the latest version LTPs123 of September 2015，there appeared a“new”species named Chloracidobacterium thermophilum whose whole genome was released in 2007 under the name Candidatus Chloracidobacterium thermophilum B［33］. The Candidatus status has prevented it from being included into the previous releases of LVTree until EMBL made a change recently. However，in LVTree C. thermophilum gets immersed in the phylum Elusimicrobia in contradiction to the original taxonomic assignment to phylum Acidobacteria. As CVTree［8］also clearly supports its belonging to Acidobacteria，it is suspected that a wrong sequence has been used in LVTree. Indeed，the GenBank accession number CP002514 cited in LVTree corresponds to a whole genome instead of a 16S rRNA. Hopefully，this point will be clarified in the next release of LVTree.

Given 11，939 of 16S rRNA sequences in total，it is remarkable that so few lineage modifications，caused either by sequence error or by taxonomic misplacement，are capable to bring all but two prokaryotic phyla to monophyletic clusters. Furthermore，it is not surprising that most of the suggested modifications are introduced for taxa proposed in the prerRNA era. Some modifications simply hint on necessary taxonomic revisions according to modern 16S rRNA sequence analysis. However，we refrain from making formal emendation for the time being in order to wait for more phenotypic and phylogenetic support.

Figure 5　An appropriately-collapsed LVTree based on all 11，939 16S rRNA sequences

Phylogeny and taxonomy at the phylum level

The aforementioned lineage modifications have added 11 monophyletic phyla. These include Acidobacteria｛23+2｝，Actinobacteria｛2892+5｝，Bacteroidetes｛1229+11｝，Caldithrixae｛2｝，Chlorobi｛11｝，Deferribacteres｛11｝，Elusimicrobia ｛1+5｝，F(xiàn)usobacteria｛39｝，Ignavibacteriae｛2｝，Synergistetes ｛23｝，and Tenericutes｛186｝. Only two phyla，F(xiàn)irmicutes｛2090 +134｝andProteobacteria｛4297+116｝，remainnonmonophyletic. Being“big”divisions，the last two phyla accommodate 56%of the 16S rRNA sequences in the LVTree dataset. In addition，a single species without proper phylum assignment（Thermoanaerobaculum aquaticum）takes the position of a separate phylum. Figure 5 shows the whole LVTree in an appropriately-collapsed form. This tree may be further scrutinized in three steps.

The first step deals with the upper 7 branches，which include 8 well-defined phyla（3 archaeal phyla collapsed into one leaf and 5 bacterial phyla）and a single family＜F＞Thermodesulfobiaceae｛3｝taking the status of phylum. Only three species are listed in this family in LPSN［10］and Vol. 3，the Bergey’s Manual［9］. These include Thermodesulfobium narugense，Coprothermobacter proteolyticus，and Coprothermobacter platensis. A 16S rRNA survey was performed in 2004［34］and recognized Coprothermobacter as an“established phylum”. Furthermore，“T. narugense is sufficiently isolated from previously described bacteria to justify creation of a new phylum”（Vol. 3，the Bergey’s Manual，page 1269）. The whole-genome-basedCVTreeanalysisalsosupports Coprothermobacter to acquire the status of a separate phylum ［35］.

＜D＞Bacteria（2282/11156+143）makes the second step of investigation which invokes the most taxonomically-confused branch. The third step starts by expanding the last branch＜D＞Bacteria｛8775/11156+184｝to a monophyletic branch made of 21 well-defined phyla，in addition to a split phylum Proteobacteria，and a single species T. aquaticum（Figure 4）. The last species takes the status of a new phylum，though the original report［36］puts it in Acidobacteria Subdivision 23. Although the same 16S rRNA sequence（GenBank accession No. JX420244）is used in［36］and in LVTree，the conclusion is clearly different. Since the whole-genome-based CVTree approach［8］puts Thermoanaerobaculum aquaticus in the phylum Acidobacteria［37］in agreement with［36］，the LVTree placement of this species requires further examination.

Traditionally，the phylum Proteobacteria contains five classes，from Alphaproteobacteria to Epsilonproteobacteria. In recent years，a few new classes have been proposed，for example，Zetaproteobacteria［38］，Acidithiobacillia［39］，and Oligoflexia［40，41］. In LVTree，instead of forming a monophyletic branch，Proteobacteria splits into two disjointed clusters.

The first cluster contains the Alpha-，Beta-，Gamma-，and Zeta-proteobacteria，as well as the class Acidithiobacillia，representing altogether 4010 rRNA sequences. A few lineage modifications would make both Alpha- and Beta- groups monophyletic. Carl Woese and coworkers observed in the mid 1990s［42］that the Beta- group gets inserted into Gamma- and these two groups，taken together，make a larger monophyletic cluster. This situation holds true if the newlyproposed class Acidothiobacillia［39］is considered a part of the Gamma-group as it was before. The second cluster comprises the Delta- and Epsilon-proteobacteria，as well as the newly-proposed Oligoflexia classes，representing 396 rRNA sequences in total. While＜C＞Epsilonproteobacteria｛111｝comes out as a monophyletic group from the outset，all species in the Delta-group appear to be intermixed. Besides the questionable class status of Oligoflexia，the whole cluster may be called a Delta/Epsilon-group.

Now we arrive at the most intricate part of Figure 5，namely，what was obtained by expanding the branch＜D＞Bacteria（2282/11156+143）. Except the three monophyletic phyla Dictyoglomi｛2｝，Cyanobacteria｛6+9｝，and Tenericutes ｛186｝，all other sequences come from the phylum Firmicutes with many taxonomic uncertainties. Actually，the branching order in this part challenges the phylum status of all the four named divisions，including Cyanobacteria whose taxonomy has been a long-due problem，partially because it has been treated under the Botanic Code and partially owning to the under-representation of 16S rRNA sequences in LVTree.

Historically，many phyla have been extracted from the phylum Firmicutes. For instance，at least four other phyla were recognized within Firmicutes［43］. Actinobacteria was an order within Firmicutes in the first edition of Bergey’s Manual，and then promoted to a class in mid 1990s，and，finally，to a separate phylum in 2012（Vol. 5，the Bergey’s Manual），whereas Tenericutes changed its class status in Firmicutes to an individual phylum in 2010（Vol. 4，the Bergey’s Manual）. Yet Firmicutes remains a warehouse for new higher-rank taxa.

In a sense，taxonomic problems of prokaryotes are now concentrated in just one line of Figure 5 as discussed in the two paragraphs above. Since the aim of this paper is to introduce a convenient tool for studying LVTree，the discussion of these classification problems would take us too far afield. However，the combined use of 16S rRNA sequence analysis by the present LVTree Viewer and the whole-genome-based CVTree web server［6-8］will surely bring prokaryotic phylogeny and taxonomy to a much better shape. However，we emphasize that all the lineage modifications suggested in this paper only serve as demonstration of how to use the interactive tree display. Any actual taxonomic revision must follow the International Code for Bacterial Nomenclature［44，45］and be published in an appropriate journal.

Authors’contributions

BH conceived the idea of the work. GZ designed and implemented the LVTree Viewer. GZ，XZ，and ZX collected sequence data of species under question. BH and XZ carried out the taxonomic comparison and suggested most of the lineage modifications. All authors read and approved the final manuscript.

Competing interests

The authors have declared that no competing interests exist.

Acknowledgments

This work was supported by the National Basic Research Program of China（973 Program；Grant No. 2013CB834100），the State Key Laboratory of Applied Surface Physics and the Department of Physics，F(xiàn)udan University，China.

Supplementary material

Supplementary material associated with this article can be found，in the online version，at http://dx.doi.org/10.1016/j. gpb.2015.12.002.

References

［1］Yarga P，Richter M，Replies J，Euze′by J，Amann R，Schleifer KH，et al. The All-Species Living Tree project: a 16S rRNA based phylogenetic tree of all sequenced type strains. Syst Appl Microbiol 2008；31:241-50.

［2］Yarga P，Ludwig W，Euze′by J，Amann R，Schleifer KH，Glo¨ckner FO，et al. Update of the All-Species Living Tree project based on 16S and 23S rRNA sequence analyses. Syst Appl Microbiol 2010；33:291-9.

［3］Munoz R，Yarza P，Ludwig W，Euze′by JP，Amann A，Schleifer KH，et al. Release LTPs104 of the All-Species Living Tree. Syst Appl Microbiol 2011；34:169-70.

［4］Yarga P，Spro¨er C，Swiderski J，Mrotzek N，Spring S，Tindall BJ，et al. Sequencing orphan species initiative（SOS）:filling gaps in the 16S rRNA gene sequence database for all species with validly published names. Syst Appl Microbiol 2013；36:69-73.

［5］Qi J，Wang B，Hao BL. Whole proteome prokaryote phylogeny without sequence alignment: a K-string composition approach. J Mol Evol 2004；58:1-11.

［6］Qi J，Luo H，Hao BL. CVTree: a phylogenetic tree reconstruction tool based on whole genomes. Nucleic Acids Res 2004；32:W45-7.

［7］Xu Z，Hao BL. CVTree update: a newly designed phylogenetic study platform using composition vectors and whole genomes. Nucleic Acids Res 2009；37:W174-8.

［8］Zuo GH，Hao BL. CVTree3 web server for whole-genome-based and alignment-free prokaryotic phylogeny and taxonomy. Genomics Proteomics Bioinformatics 2015；13:321-31.

［9］Bergey’s Manual Trust. Bergey’s Manual of systematic bacteriology，2nd ed.，vols. 1-5. New York: Springer-Verlag；2001-2012.

［10］Parte AC. LPSN - list of prokaryotes with standing in nomenclature. Nucleic Acids Res 2014；42:D613-6.

［11］Lawson PA，Gharbia SE，Shah HN，Clark DR，Collins MD. Intrageneric relationships of members of the genus Fusobacterium as determined by reverse transcriptase sequencing of smallsubunit rRNA. Int J Syst Bacteriol 1991；41:347-54.

［12］Jalava J，Eerola E. Phylogenetic analysis of Fusobacterium alocis and Fusobacterium sulci based on 16S rRNA gene sequence: proposal of Filifactor alocis（Cato，Moore and Moore）comb. nov. and Eubacterium sulci（Cato，Moore and Moore）comb. nov.. Int J Syst Bacteriol 1999；49:1375-9.

［13］Hyerman J，Verbeenen J，Schumann P，Swings J，De Vos P. Six novel Arthrobacter species isolated from deteriorated mural paintings. Int J Syst Evol Microbiol 2005；55:1457-64.

［14］Soutschek E，Winter J，Schindler F，Kandler O. Acetomicrobium flavidum，gen. nov.，sp. nov.，a thermophilic，anaerobic bacterium from sewage sludge，forming acetate，CO2and H2from glucose. Syst Appl Microbiol 1984；5:377-90.

［15］Winter J，Buaun E，Zabel HP. Acetomicrobium faecale，sp. nov.，a strictly anaerobic bacterium from sewage sludge，producing ethanol from pentoses. Syst Appl Microbiol 1987；9:71-6.

［16］Mavromatis K，Stackebrandt E，Held B，Lapidus A，Nolan M，Lucas S，et al. Complete genome sequence of the moderate thermophile Anaerobaculum mobile type strain（NGAT）. Stand Genomic Sci 2013；8:47-57.

［17］Jumas-Bilak E，Roudiere L，Marchandin M. Description of ‘Synergistetes’phyl. nov. and emended description of the phylum ‘Deferribacteres’and of the family Syntrophomonadaceae，phylum ‘Firmicutes’. Int J Syst Evol Microbiol 2009；59:1028-36.

［18］Bouanane-Darenfed A，Ben Hania W，Hacene H，Cayol JL，Ollivier B，F(xiàn)ardeau ML. Caldicoprobacter guelmensis sp. nov.，a thermophilic，anaerobic，xylanolytic bacterium isolated from a hot spring. Int J Syst Evol Microbiol 2013；63:2049-53.

［19］Shah HN，Collins MD. Reclassification of Bacteroides furcosus Veillon and Zuber 1898（Handuroy，Urbain，Gillot and Magron）in a new genus Anaerorhabdus，as Anaerorhabdus furcosus comb. nov. Syst Appl Microbiol 1986；8:86-8.

［20］Horino H，F(xiàn)ujita T，Tonouchi A，et al. Description of Anaerobacterium chartisolvens gen. nov.，sp. nov.，an obligately anaerobic bacteriumfrom ClostridiumrRNAcluster IIIisolatedfromsoilofa Japanese rice field，and reclassification of Bacteroides cellulosolvens Murray，1984 as Pseudobacteroides cellulosolvens gen. nov.，comb. nov.. Int J Syst Evol Microbiol 2014；64:1296-303.

［21］Kim DS，Lee YJ，Yoon SH，Kim N，Na H，Park SC，et al. Introducing EzTaxon-e: a phylogenetic 16S rRNA gene sequence database with phylotypes that represent uncultured species. Int J Syst Evol Microbiol 2012；62:716-21.

［22］Patel NB，Tito RY，Obregon-Tito AJ，O’Neal L，Trujillo-Villavoel O，Marin-Reyes L，et al. Ezakiella peruensis gen. nov.，sp. nov.，isolated from human fecal sample from a coastal traditional community in Peru. Anaerobe 2015；32:43-8.

［23］Jensen NS，Canale-Parola E. Bacteroides pectinophilus sp. nov. and Bacteroides galacturonicus sp. nov.: two pectinolytic bacteria from the human intestinal tract. Appl Environ Microbiol 1986；52:880-7.

［24］Kim BK，Kim OS，Moon EY，Chun J. Proposal to transfer Flavobacterium oceanosedimentum Carty and Litchfield 1978 to the genus Curtobacterium as Curtobacterium oceanosedimentum comb. nov. FEMS Microbiol Lett 2009；296:137-41.

［25］Miroshnichenko ML，Kostrikina NA，Chernyh NA，Pimenov NV，Tourova TP，Antipov AN，et al. Caldithrix abyssi gen. nov.，sp. nov.，a nitrate-reducing，thermophilic，anaerobic bacterium isolated from a Mid-Atlantic Ridge hydrothermal vent，represents a novel bacterial lineage. Int J Syst Evol Microbiol 2003；53:323-9.

［26］Miroshnichenko ML，Kolganova TV，Spring S，Chernyh N，Nonch-Osmolovskaya EA. Caldithrix palaeochoryensis sp. nov.，a thermophilic，chemo-organotrophic bacterium from a geothermally heated sediment，and emended description of the genus Caldithrix. Int J Syst Evol Microbiol 2010；60:2120-3.

［27］Collins MD，Lawson PA，Willems A，Cordoba JJ，F(xiàn)ernandez-Garayzabel J，Garcia P，et al. The phylogeny of the genus Clostridium: proposal of five new genera and eleven new species combinations. Int J Syst Bacteriol 1994；44:812-26.

［28］Grossling VBD. Moore WEC. Gemmiger formicilis，n. gen.，n. sp.，an anaerobic budding bacterium from intestine. Int J Syst Bacteriol 1975；25:202-7.

［29］Gromov BV，Mamkayeva KA. Electron microscope study of parasitism by Bdellovibrio chlorellavorus on cells of the green algae Chlorella vulgaris. Tsitologia 1972；14:256-60.

［30］Gromov BV，Mamkayeva KA. Proposal of a new genus Vampirovibrio for chlorellavorus bacteria previously assigned to Bdellovibrio. Mikrobiologia 1980；49:165-7.

［31］Robinson IM，F(xiàn)reundt EA. Proposal for an amended classification of anaerobic mollicutes. Int J Syst Bacteriol 1987；37:78-81.

［32］Antunes A，Rainey FA，Wanner G，Taborda M，Patzold J，F(xiàn)ernanda Nobre M，et al. A new lineage of halophilic，wall-less，contractile bacteria from a brine-filled deep of the Red Sea. J Bacteriol 2008；190:3580-7.

［33］Bryant DA，Costas AM，Maresca JA，Chew AG，Klatt CG，Bateson MM，et al. Candidatus Chloracidobacterium thermophilum: an aerobic phototrophic Acidobacterium. Science 2007；317:523-6.

［34］Schloss PD，Handelsman J. Status of the microbial census. Microbiol Mol Biol Rev 2004；68:686-91.

［35］Hao BL. CVTrees support the Bergey’s systematics and provide high resolution at species level and below. Bull BISMiS 2011；2:189-96.

［36］Losey NA，Stevenson BS，Busse HJ，Sinninghe Damste JS，Rijpstra WI，Rudd S，et al. Thermoanaerobaculum aquaticus gen. nov.，sp. nov.，the first cultivated member of Acidobacteria subdivision 23，isolated from hot spring. Int J Syst Evol Microbiol 2013；63:4149-57.

［37］Stamps BW，Losey NA，Lawson PA，Stevenson BS. Genome sequence of Thermoanaerobaculum aquaticum MP-01T，the first cultivated member of Acidobacteria subdivision 23，isolated from a hot spring. Genome Announc 2014；2，e00570-14.

［38］Emerson D，Rentz JA，Lilburn TG，Davis RE，Aldrich H，Chan C，et al. A novel lineage of Proteobacteria involved in formation of marine Fe-oxidizing microbial mat communities. PLoS One 2007；2:e667.

［39］Williams KP，Kelley DP. Proposal for a new class within the phylum Proteobacteria，Acidithiobacillia classis nov.，with the type order Acidithiobacillales，and emended description of the class Gammaproteobacteria. Int J Syst Evol Microbiol 2013；63:2901-6.

［40］Nakai R，Nishijima M，Tazato N，Handa Y，Karray F，Sayada S，et al. Oligoflexus tunisiensis gen. nov.，sp. nov.，a Gram-negative，aerobic，filamentous bacterium of a novel proteobacterial lineage，and description of Oligoflexaceae fam. nov.，Oligoflexales ord. nov. and Oligoflexia classis nov. Int J Syst Evol Microbiol 2014；64:3353-9.

［41］Nakai R，Nagamura T. Oligoflexia，the newest class of the phylum Proteobacteria，consisting of only one cultured species and uncultured bacterial phylotypes in diverse habitats. J Phylogen Evol Biol 2015；3:141.

［42］Olsen GJ，Woese CR，Overbeek R. The winds of（evolutionary）change: breathing new life into microbiology. J Bacteriol 1994；176:1-6.

［43］Garrity GM，Bell JA，Lilburn TG. Taxonomic outline of prokaryotes. Bergey’s manual of systematic bacteriology Release 5.0. 2nd ed. New York，Berlin，Heidelberg: Springer；2004. http:// dx.doi.org/10.1007/bergeysoutline200405.

［44］Lapage SP，Sneath PHA，Lessal EF，Skerman VBD，Seeliger HPR，Clark WA，editors. International code of nomenclature of bacteria（1990 Revision）. Washington，DC: American Society for Microbiology；1992.

［45］De Vos P，Truper HG. Judicial Commission of the International Committee on Systematic Bacteriology，IXth International （IUMS）Congress of Bacteriology and Applied Microbiology，Minutes of the meetings，14，15 and 18 August 1999，Sydney，Australia. Int J Syst Evol Microbiol 2000；50:2239-44.

METHOD

*Corresponding author.