VIRGO2: Unveiling the Functional and Ecological Complexity of the Vaginal Microbiome with an Enhanced Non-Redundant Gene Catalog

VIRGO2: Unveiling the Functional and Ecological Complexity of the Vaginal Microbiome with an Enhanced Non-Redundant Gene Catalog | bioRxiv /* */ /* */ <!-- <!-- /*! * yepnope1.5.4 * (c) WTFPL, GPLv2 */ (function(a,b,c){function d(a){return"[object Function]"==o.call(a)}function e(a){return"string"==typeof a}function f(){}function g(a){return!a||"loaded"==a||"complete"==a||"uninitialized"==a}function h(){var a=p.shift();q=1,a?a.t?m(function(){("c"==a.t?B.injectCss:B.injectJs)(a.s,0,a.a,a.x,a.e,1)},0):(a(),h()):q=0}function i(a,c,d,e,f,i,j){function k(b){if(!o&&g(l.readyState)&&(u.r=o=1,!q&&h(),l.onload=l.onreadystatechange=null,b)){"img"!=a&&m(function(){t.removeChild(l)},50);for(var d in y[c])y[c].hasOwnProperty(d)&&y[c][d].onload()}}var j=j||B.errorTimeout,l=b.createElement(a),o=0,r=0,u={t:d,s:c,e:f,a:i,x:j};1===y[c]&&(r=1,y[c]=[]),"object"==a?l.data=c:(l.src=c,l.type=a),l.width=l.height="0",l.onerror=l.onload=l.onreadystatechange=function(){k.call(this,r)},p.splice(e,0,u),"img"!=a&&(r||2===y[c]?(t.insertBefore(l,s?null:n),m(k,j)):y[c].push(l))}function j(a,b,c,d,f){return q=0,b=b||"j",e(a)?i("c"==b?v:u,a,b,this.i++,c,d,f):(p.splice(this.i++,0,a),1==p.length&&h()),this}function k(){var a=B;return a.loader={load:j,i:0},a}var l=b.documentElement,m=a.setTimeout,n=b.getElementsByTagName("script")[0],o={}.toString,p=[],q=0,r="MozAppearance"in l.style,s=r&&!!b.createRange().compareNode,t=s?l:n.parentNode,l=a.opera&&"[object Opera]"==o.call(a.opera),l=!!b.attachEvent&&!l,u=r?"object":l?"script":"img",v=l?"script":u,w=Array.isArray||function(a){return"[object Array]"==o.call(a)},x=[],y={},z={timeout:function(a,b){return b.length&&(a.timeout=b[0]),a}},A,B;B=function(a){function b(a){var a=a.split("!"),b=x.length,c=a.pop(),d=a.length,c={url:c,origUrl:c,prefixes:a},e,f,g;for(f=0;f<d;f++)g=a[f].split("="),(e=z[g.shift()])&&(c=e(c,g));for(f=0;f<b;f++)c=x[f](c);return c}function g(a,e,f,g,h){var i=b(a),j=i.autoCallback;i.url.split(".").pop().split("?").shift(),i.bypass||(e&&(e=d(e)?e:e[a]||e[g]||e[a.split("/").pop().split("?")[0]]),i.instead?i.instead(a,e,f,g,h):(y[i.url]?i.noexec=!0:y[i.url]=1,f.load(i.url,i.forceCSS||!i.forceJS&&"css"==i.url.split(".").pop().split("?").shift()?"c":c,i.noexec,i.attrs,i.timeout),(d(e)||d(j))&&f.load(function(){k(),e&&e(i.origUrl,h,g),j&&j(i.origUrl,h,g),y[i.url]=2})))}function h(a,b){function c(a,c){if(a){if(e(a))c||(j=function(){var a=[].slice.call(arguments);k.apply(this,a),l()}),g(a,j,b,0,h);else if(Object(a)===a)for(n in m=function(){var b=0,c;for(c in a)a.hasOwnProperty(c)&&b++;return b}(),a)a.hasOwnProperty(n)&&(!c&&!--m&&(d(j)?j=function(){var a=[].slice.call(arguments);k.apply(this,a),l()}:j[n]=function(a){return function(){var b=[].slice.call(arguments);a&&a.apply(this,b),l()}}(k[n])),g(a[n],j,b,n,h))}else!c&&l()}var h=!!a.test,i=a.load||a.both,j=a.callback||f,k=j,l=a.complete||f,m,n;c(h?a.yep:a.nope,!!i),i&&c(i)}var i,j,l=this.yepnope.loader;if(e(a))g(a,0,l,0);else if(w(a))for(i=0;i (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start':new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0];var j=d.createElement(s);var dl=l!='dataLayer'?'&l='+l:'';j.src='//www.googletagmanager.com/gtm.js?id='+i+dl;j.type='text/javascript';j.async=true;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-M677548'); Skip to main content Home About Submit ALERTS / RSS Search for this keyword Advanced Search New Results VIRGO2: Unveiling the Functional and Ecological Complexity of the Vaginal Microbiome with an Enhanced Non-Redundant Gene Catalog View ORCID Profile M. T. France , View ORCID Profile I. Chaudry , L. Rutt , M. Quain , B. Shirtliff , E. McComb , A. Maros , M. Alizadeh , F. A. Hussain , M. A. Elovitz , View ORCID Profile D. A. Relman , A. Rahman , R. M. Brotman , J. Price , M. Kassaro , View ORCID Profile J. B. Holm , View ORCID Profile B. Ma , View ORCID Profile J. Ravel doi: https://doi.org/10.1101/2025.03.04.641479 M. T. France 1 Center for Advanced Microbiome Research and Innovation, Institute for Genome Sciences, University of Maryland School of Medicine , Baltimore, MD USA 2 Department of Microbiology and Immunology, University of Maryland School of Medicine , Baltimore, MD, USA Find this author on Google Scholar Find this author on PubMed Search for this author on this site ORCID record for M. T. France I. Chaudry 1 Center for Advanced Microbiome Research and Innovation, Institute for Genome Sciences, University of Maryland School of Medicine , Baltimore, MD USA Find this author on Google Scholar Find this author on PubMed Search for this author on this site ORCID record for I. Chaudry L. Rutt 1 Center for Advanced Microbiome Research and Innovation, Institute for Genome Sciences, University of Maryland School of Medicine , Baltimore, MD USA Find this author on Google Scholar Find this author on PubMed Search for this author on this site M. Quain 1 Center for Advanced Microbiome Research and Innovation, Institute for Genome Sciences, University of Maryland School of Medicine , Baltimore, MD USA Find this author on Google Scholar Find this author on PubMed Search for this author on this site B. Shirtliff 1 Center for Advanced Microbiome Research and Innovation, Institute for Genome Sciences, University of Maryland School of Medicine , Baltimore, MD USA Find this author on Google Scholar Find this author on PubMed Search for this author on this site E. McComb 1 Center for Advanced Microbiome Research and Innovation, Institute for Genome Sciences, University of Maryland School of Medicine , Baltimore, MD USA Find this author on Google Scholar Find this author on PubMed Search for this author on this site A. Maros 1 Center for Advanced Microbiome Research and Innovation, Institute for Genome Sciences, University of Maryland School of Medicine , Baltimore, MD USA Find this author on Google Scholar Find this author on PubMed Search for this author on this site M. Alizadeh 1 Center for Advanced Microbiome Research and Innovation, Institute for Genome Sciences, University of Maryland School of Medicine , Baltimore, MD USA Find this author on Google Scholar Find this author on PubMed Search for this author on this site F. A. Hussain 3 Department of Biology, School of Arts and Sciences, Tufts University , Medford MA Find this author on Google Scholar Find this author on PubMed Search for this author on this site M. A. Elovitz 4 Women’s Biomedical Research Institute, Icahn School of Medicine at Mount Sinai , New York, NY, USA 5 Department of Obstetrics, Gynecology and Reproductive Sciences, Icahn School of Medicine at Mount Sinai , New York, NY, USA Find this author on Google Scholar Find this author on PubMed Search for this author on this site D. A. Relman 6 Department of Medicine, Stanford University School of Medicine , Stanford, CA, USA 7 Department of Microbiology & Immunology, Stanford University School of Medicine , Stanford, CA, USA 8 Infectious Diseases Section, Veterans Affairs, Palo Alto Health Care System , Palo Alto, CA, USA Find this author on Google Scholar Find this author on PubMed Search for this author on this site ORCID record for D. A. Relman A. Rahman 9 Maternal and Child Health Division, International Centre for Diarrhoeal Disease Research , Bangladesh , Dhaka, Bangladesh Find this author on Google Scholar Find this author on PubMed Search for this author on this site R. M. Brotman 1 Center for Advanced Microbiome Research and Innovation, Institute for Genome Sciences, University of Maryland School of Medicine , Baltimore, MD USA 10 Department of Epidemiology and Public Health, University of Maryland School of Medicine , Baltimore, MD, USA Find this author on Google Scholar Find this author on PubMed Search for this author on this site J. Price 11 Department of Obstetrics and Gynecology, University of North Carolina at Chapel Hill , Chapel Hill, North Carolina, USA 12 UNC Global Projects–Zambia , Lusaka, Zambia Find this author on Google Scholar Find this author on PubMed Search for this author on this site M. Kassaro 11 Department of Obstetrics and Gynecology, University of North Carolina at Chapel Hill , Chapel Hill, North Carolina, USA 12 UNC Global Projects–Zambia , Lusaka, Zambia Find this author on Google Scholar Find this author on PubMed Search for this author on this site J. B. Holm 1 Center for Advanced Microbiome Research and Innovation, Institute for Genome Sciences, University of Maryland School of Medicine , Baltimore, MD USA 2 Department of Microbiology and Immunology, University of Maryland School of Medicine , Baltimore, MD, USA Find this author on Google Scholar Find this author on PubMed Search for this author on this site ORCID record for J. B. Holm B. Ma 1 Center for Advanced Microbiome Research and Innovation, Institute for Genome Sciences, University of Maryland School of Medicine , Baltimore, MD USA 2 Department of Microbiology and Immunology, University of Maryland School of Medicine , Baltimore, MD, USA Find this author on Google Scholar Find this author on PubMed Search for this author on this site ORCID record for B. Ma J. Ravel 1 Center for Advanced Microbiome Research and Innovation, Institute for Genome Sciences, University of Maryland School of Medicine , Baltimore, MD USA 2 Department of Microbiology and Immunology, University of Maryland School of Medicine , Baltimore, MD, USA Find this author on Google Scholar Find this author on PubMed Search for this author on this site ORCID record for J. Ravel For correspondence: jravel{at}som.umaryland.edu Abstract Full Text Info/History Metrics Supplementary material Data/Code Preview PDF Abstract Despite the importance of the cervicovaginal microbiome, the mechanisms that govern its composition and drive its impact on host physiology remain poorly understood. This study expands our understanding of the function and ecology of the vaginal microbiome using VIRGO2, an enhanced non-redundant gene catalog comprising over 1.7 million well-annotated genes from body-site specific microbes and viruses. Analyses using VIRGO2 revealed novel insights, including the identification of previously uncharacterized vaginal bacteria, features of the vaginal mycobiome and phageome, and differential expression of bacterial carbohydrate catabolic genes. Constructed from over 2,500 metagenomes and 4,000 bacterial genomes, VIRGO2 broadens geographic representation and microbial diversity compared to its predecessor. This updated catalog enables more precise profiling of taxonomic and functional composition from metagenomic and metatranscriptomic datasets. VIRGO2 is a critical resource for integrative analyses of vaginal microbial communities and their interactions with host tissues, thereby enhancing our mechanistic understanding of vaginal health and disease. Introduction Human microbiome science has progressed from identifying the resident microbes associated with health to characterizing how these microbes interact as communities within the host and contribute to host physiology. These microbial communities live in close association with host tissues and can affect physiologic change via direct interactions 1 , 2 or through the production of metabolites and secreted products 3 – 5 . A mechanistic understanding of the interactions between host and microbes can enable the manipulation of this system to improve an individual’s health 6 – 9 . In pursuit of this mechanistic understanding, shotgun metagenomic and metatranscriptomic data are often used to describe the genomic content and in vivo transcriptional activity of the communities, respectively 10 , 11 . These techniques generate sequence data from bulk extracted DNA or RNA to provide an untargeted assessment of the genomes and transcriptomes that comprise microbial communities. However, the analysis and integration of these data is no trivial task. Sequences originating from tens to hundreds of microbial species might be found in each metagenome and metatranscriptome depending on the complexity of the underlying community. Reference-guided approaches that match sequence reads to genomic databases provide an attractive strategy to simplify the analysis of these complex data 12 . Yet, these approaches can yield unsatisfactory results when the reference database poorly matches the microbiome being analyzed 13 . Ecosystem-specific databases that are tailored to a particular habitat and microbial community can provide a more robust analysis of metagenomic and metatranscriptomic data 14 and have been published for several different host site-specific microbiomes 15 – 19 . Here, we present VIRGO2, an enhanced iteration of the non-redundant gene catalog for the human vaginal microbiome, VIRGO (hereafter referred to as VIRGO1) 20 , and demonstrate its use in gaining functional and ecological insight into this important microbiome with the re-analysis of 2,560 metagenomes and 363 metatranscriptomes. The microbial communities that inhabit the human vagina are distinct from those found at other body sites, where community diversity roughly correlates with health 21 . Instead, an optimal vaginal microbiota (VMB) is often characterized by dominance of a single species of Lactobacillus and is associated with decreased risk for several adverse health outcomes (reviewed in France, et al. 22 ) including the acquisition of sexually transmitted infections 23 – 25 , spontaneous preterm birth 26 , 27 , and pelvic-inflammatory disease 28 , 29 . Lactobacilli are thought to dampen inflammatory responses 24 , 30 , 31 and to exclude vaginal pathogens by lowering vaginal pH through their production of lactic acid 32 . A non-optimal VMB, on the other hand, is comprised of a more diverse and even mixture of facultative and obligate anaerobic bacterial species in the genera Gardnerella , Prevotella , Fannyhessea , Sneathia , Ca. Lachnocurva vaginae and Megasphaera , among others 33 . Women with these communities are often diagnosed with bacterial vaginosis, a common vaginal condition clinically defined by an elevated pH with a concomitant thin and homogenous discharge, fishy odor, and the presence of clue cells 34 . Yet, many women with a non-optimal VMB do not report experiencing any of these signs and symptoms 35 , but still appear to be at risk 36 – 38 . The composition of the VMB can also change over time, including shifting from an optimal Lactobacillus -dominated state to a non-optimal state, or the reverse 39 – 41 . Understanding the factors that govern the composition and functions of the VMB, and the mechanisms by which vaginal microbes influence host tissues, will enable the development of novel therapeutics and treatment options to promote vaginal health. The VMB can be subdivided by taxonomic composition into seven community state types (CSTs) 42 , four of which are dominated by lactobacilli: I- L. crispatus , II- L. gasseri / L. paragasseri , III- L. iners , V- L. jensenii / L. mulieris . The remaining three CSTs (IV-A, IV-B, IV-C) are characterized by a diverse mixture of facultative and obligate anaerobes; IV-A is distinguished by the presence of Ca . L. vaginae (formerly BVAB1), while IV-B is characterized by higher relative abundance of Gardnerella spp. and IV-C by the absence of Gardnerella and Fannyhessea . Of the four Lactobacillus -dominant CSTs, only CST I, II, and V are considered optimal. L. iners -dominated communities shift to CST IV at greater rates than those dominated by other lactobacilli, and do not appear to offer the same degree of protective benefits to the host 39 , 43 – 46 . Recent studies of the VMB have utilized shotgun metagenomics and metatranscriptomics to describe the vaginal microbiome and have revealed additional complexity not captured by the CST paradigm. For example, in the release of VIRGO1, we highlighted that vaginal bacterial populations appear to not be genetically homogenous, and instead comprised of multiple genotypes 20 . Since then, we have leveraged this intraspecies diversity to identify more complex microbial population configurations that are common across women and are associated with symptomatic BV 47 . Other studies have found that genetic diversity within Gardnerella is associated with preterm birth 27 , 48 . Analyses of the vaginal metatranscriptome revealed further complexity by demonstrating that the transcriptional activity of a bacterium is not always correlated with its abundance in a community 49 , 50 . Still more studies have integrated metagenomic data with metabolomic data to identify and characterize enzymatic differences between L. iners and other vaginal lactobacilli 51 , 52 . From all these studies, it has become clear that integrative multi-omic analyses are needed to fully understand the vaginal microbiome and its relationship to health. Here, we describe the construction, validation and application of VIRGO2, an enhanced iteration of a non-redundant gene (NRG) catalog for the vaginal microbiome. Since its release, VIRGO1 has seen regular use in the analysis of vaginal metagenomic and metatranscriptomic data and has proven to be a powerful tool for analyzing and integrating these data types. VIRGO2 was constructed ab initio from a large sequence dataset of >2,500 shotgun metagenomes and >4,000 whole genome sequences from relevant bacteria, fungi, viruses, and protists. For this version, we emphasized broadening the geographic representation of the hosts from which sequence data were collected to build the catalog and improving VIRGO’s ability to characterize the non-bacterial components of the vaginal microbiome. The resulting NRG catalog is approximately twice the size of the original and was validated for its ability to characterize the vaginal microbiome. A re-analysis of vaginal metagenomes and metatranscriptomes demonstrated how VIRGO2 can be used to gain insight into the VMB and highlighted novel ecological and functional findings related to the bacteria, bacteriophages, and fungi that inhabit the human vagina. Results De novo assembly and binning of shotgun metagenomes to generate MAGs The majority of the sequence data used to construct VIRGO2 was sourced from 2,560 shotgun metagenomes derived from vaginal (n=2,496), or penile urethral (n=64) samples (see Supplementary Table 1 for accession numbers; metagenomes originate from previously published studies 23 , 50 , 53 – 62 ). The small collection of penile metagenomes was derived from a North American sexual network study 55 and was included to more comprehensively characterize the organisms that might be identified in vaginal ‘omics data. In total, the metagenomes were derived from samples provided by 2,048 individuals located in six different countries and representing five different continents ( Fig. 1a ). All samples were provided by reproductive-age individuals, and 1,173 (47%) of the vaginal metagenomes were derived from samples provided by pregnant women. The metagenomes from North America derived from pregnant and non-pregnant women. All other continents were represented by samples from either only non-pregnant women (Fiji) or only pregnant women (Europe, Asia, Africa). Download figure Open in new tab Fig. 1 Overview of the metagenomes, metagenome-assembled genomes, and bioinformatics processes used to construct VIRGO2. a, Geographic origin of shotgun metagenomes used in the construction of VIRGO2 (see Supplementary Table 1 for accession numbers). b, Bioinformatics pipeline used to generate metagenome assembled genomes from the shotgun metagenomes. c, CheckM completeness and d, contamination estimates of the 15,136 MAGs generated in this study. e, Association between the number of MAGs recovered and the community state type of the metagenome. Numbers above the violin plots denote the number of metagenomes assigned to each CST. Each metagenome was processed using a de novo assembly and binning pipeline ( Fig. 1b ) to generate metagenome assembled genomes (MAGs). This pipeline produced a total of 10,467 high and 4,670 medium quality MAGs, of which 10,598 (70.0%) were estimated to be more than 90% complete ( Fig. 1c ) and 14,591 (96.4%) were estimated to be less than 5% contaminated ( Fig. 1d ). Many of medium quality MAGs that were estimated to have lower completeness represented poorly characterized species (Supplementary Table 2, e.g. Ca . L. vaginae, average completeness 75.8%). Almost all the metagenomes produced at least one MAG (n=2,412; 94.2%), and many produced several (range: 0-48, median: 3). Not surprisingly, the number of MAGs produced from a metagenome was found to vary with community state type. Metagenomes that were dominated by single species yielded fewer MAGs than those that were more compositionally diverse and even ( Fig. 1e ; e.g. CST I-A vs IV-B, Poisson regression, p adj < 0.01). A comprehensive inventory of the bacteria that inhabit the human vagina The collection of 15,137 MAGs represents a broad and unbiased sampling of the bacterial species which inhabit the vaginal microbiome. The taxonomy of the MAG collection was established using the Genome Taxonomy Database Toolkit (GTDB-tk), revealing 288 unique bacterial species, including 146 that were present in at least five metagenomes. Among these, Lactobacillus iners was the most prevalent species identified and was found in 1,380 metagenomes (53.9%, Fig. 2a ). L. iners MAGs were routinely produced from communities where it was the dominant species, but also from communities where it was co-resident with other lactobacilli or with non- Lactobacillus species ( Fig. 2a , radar plot). Other familiar prevalent species include F. vaginae (formerly Atopobium vaginae , 27.6%), M. lornae (25.9%), L. crispatus (24.1%), Amygdalobacter indicium (formerly BVAB2 63 , 19.6%), Ca. L. vaginae (formerly BVAB1, 15.1%), as well as several species in the genera Gardnerella , Prevotella , and Fannyhessea ( Fig. 2a ). Our analysis also revealed some prevalent species that, although recognized in the GTDB, have yet to be given names, including: Berryella sp001552935 (21.8%), Dialister sp001553355 (15.9%), Parvimonas sp001552895 (11.9%), and Prevotella sp000758925 (9.3%). All four of these bacteria were essentially only identified in communities that were not dominated by Lactobacillus ( Fig. 2a , radar plots). Download figure Open in new tab Fig. 2 Description of the taxonomy and prevalence of vaginal bacteria. a, Treemap diagram displaying the number of MAGs generated for each of the 50 most prevalent species. Empty spaces within phyla represent the number of residual MAGs in that phylum not accounted for by the displayed species. Interior radar plots indicate the CST distribution of the metagenomes that originated the species’ MAGs. Concatenated core genome phylogenetic trees of the genera where prominent novel species were identified including: b, Gardnerella, c, Fannyhessea and d, Prevotella . Leaves are labeled with the species’ designation along with the species’ average genome size and the number of MAGs generated. The collection of MAGs also included many genomes which could not be assigned to a species by the GTDB-tk, indicating they may belong to novel or poorly characterized species. To classify these genomes, we first collected representative genomes of closely related species and then performed a phylogenetic analysis. Two novel species were identified in the genus Gardnerella , increasing the total number of Gardnerella species to 15 64 . Neither novel Gardnerella species were particularly prevalent, being identified in just 3 (0.1%) and 40 (1.6%) metagenomes respectively ( Fig. 2b ). We have since isolated the second of these novel Gardnerella species from a vaginal swab specimen and confirmed its identity via whole genome sequencing (ascn: ACNEUN). Three novel species were identified in Fannyhessea ( Fig. 2c ), two that were rare ( Fannyhessea sp. novel 1, n=1; Fannyhessea sp. novel 2, n=2), and one that was common ( Fannyhessea sp. novel 3, n=358, 14.0%). We found Fannyhessea. sp. novel 3 to share a clade with Fannyhessea sp. novel 1 whereas, Fannyhessea sp. novel 2 was more closely related to F. vaginae and F. massiliensis ( Fig. 2c ). Another three novel species were identified in Prevotella , all of which were fairly common. Prevotella sp. novel 1 was identified in 148 metagenomes (5.8%) and is related to P. ihumii ( Fig. 2d ). The remaining two novel Prevotella species share a clade with P. timonensis and P. sp000479005, and of these four, Prevotella sp. novel 3 was the most frequently identified (n=343, 13.3%), surpassing even P. timonensis (n=314, 12.2%). Construction and annotation of the VIRGO2 non-redundant gene catalog To construct the VIRGO2 gene catalog, the 15,136 MAGs were supplemented with sequence data from: 1) the residual un-binned metagenome contigs (length >2,000bp) and 2) isolate whole genome sequences (WGS) from relevant bacteria, viruses, fungi, and protists (254 unique species, 4,013 genomes, accession numbers in Supplementary Table 3). Prodigal identified 21,202,535 coding sequences (CDS) from the MAGs, 8,294,162 from the residual un-binned contigs and 8,015,873 from the isolate WGS (total=37,512,570 CDS). The CDS were then clustered using CDhit (95% sequence identity and 90% overlap), and for each cluster, the longest gene was selected as a representative, producing 1,773,155 non-redundant genes (NRGs), roughly a doubling of that included in the VIRGO1 gene catalog (∼1.8M versus ∼0.9M). Around one third of the NRGs were unique (cluster size of 1, n=673,407) and the remaining two thirds had a median cluster size of 5 (Supplementary Fig. 1). The NRGs were next annotated with taxonomic and functional information to facilitate their use in characterizing vaginal microbiome ‘omics data. The taxonomic annotation of each NRG was derived from that of the source sequences in the corresponding gene cluster ( Fig. 3a ). In total, taxonomic annotations were applied to 91% of the NRGs in VIRGO2, which is a dramatic improvement over VIRGO1 (47%). Functional annotations were also applied to the NRG using eggNOG-mapper 65 (v2, default settings), PANNZER2 66 , and AMRFinderPlus 67 . The applied functional schemes included eggNOG 68 , COG category 69 , PFAM 70 , GO terms 71 , Enzyme Commission numbers 72 , BiGG Reaction 73 , KEGG 74 , and Carbohydrate Active Enzymes 75 , Gene Product 66 , antimicrobial resistance genes 67 , and secondary metabolite biosynthetic gene pathways 76 . Combined, 81% of the VIRGO2 NRGs have at least one functional annotation (Supplementary Fig. 2). Download figure Open in new tab Fig. 3 Annotation and validation of the VIRGO2 non-redundant gene catalog. a, Sankey diagram displaying the taxonomic assignment decision tree for VIRGO2 non-redundant genes. b, Gene accumulation curves showing the change in the number of non-redundant genes for a species as the number of metagenomes containing that species increases. Inset graph shows the accumulation curve for the overall number of non-redundant genes in VIRGO2. c, Comparison of the expected (left) and VIRGO2 estimated (right) taxonomic compositions of mock shotgun metagenomes. d, Relationship between expected and observed relative abundances (linear regression, r 2 =.988, p<0.001). Arrows indicate instances where VIRGO2 is slightly underestimating the relative abundances of G. swidsinskii and L. paragasseri . Validation of VIRGO2 demonstrates comprehensiveness and accuracy VIRGO2 is intended to comprise a union of pangenomes of the microbes that inhabit the vagina, i.e., the vaginal microbiome’s “meta-pangenome” 77 . To test if VIRGO2 adequately encompasses each species’ genetic diversity, we generated NRG accumulation curves that track the total number of NRGs with increasing numbers of metagenomes. For the catalog overall ( Fig. 3b , inlay), and for most of the individual species we examined, the curves demonstrated a rapid saturation in the repertoire of NRGs ( Fig. 3b ). However, for some species the accumulation curves failed to approach an asymptote, indicating that VIRGO2 may benefit from the inclusion of data from more samples containing these species. Among those not reaching saturation were L. crispatus and P. timonensis , two prominent vaginal bacteria that were well-represented in the metagenomes used to build VIRGO2. To define the scope of genetic diversity in these two species, we grouped their NRGs into orthologous gene clusters and generated pangenome accumulation curves. For both species, the accumulation curves approach saturation at ∼7000 orthologs (Supplementary Fig. 3), indicating that much of their diversity in NRG content originates from genetic variants of orthologous genes. These results demonstrate that VIRGO2 comprises the genetic and functional diversity of vaginal microbes. We next sought to validate VIRGO2’s ability to characterize the taxonomic composition of vaginal microbiome ‘omics data. To do this, we constructed mock communities containing known mixtures of species using synthetic metagenomic sequence reads. Some of the mock communities were designed to mimic natural community structures, while others comprised random assortments, or specific combinations of closely related species (e.g. L. jensenii and L. mulieris , Fig. 3c ). The synthetic reads were mapped to the VIRGO2 NRG catalog and taxonomy was estimated as relative abundances (corrected for gene length). We found that VIRGO2 accurately characterized community composition even when presented with particularly challenging cases ( Fig. 3c ). For example, VIRGO2 was able to disentangle the relative abundances of closely related Gardnerella species when presented with simulated mixtures of varying proportions. One exception was that VIRGO2 appears to slightly underestimate the relative abundances of G. swidsinskii and L. paragasseri ( Fig. 3d , arrows). Further investigation revealed that both species have close relatives ( G. leopoldii and L. gasseri , respectively) that share some NRG clusters, causing them to be labeled at the genus, rather than species level. Nevertheless, the overall correlation between expected and observed relative abundances for these synthetic mock datasets was high (r 2 0.988, p<0.001, linear regression, Fig. 3d ), indicating that VIRGO2 accurately characterizes vaginal microbial community composition. Taxonomic composition of the vaginal microbiota as characterized by VIRGO2 One of the main functions of VIRGO2 is to estimate the composition of vaginal microbiome ‘omics datasets. To demonstrate this, we characterized the taxonomic composition of the 2,560 metagenomes used in the construction of the catalog. Sequence reads were mapped to VIRGO2, and relative abundances were estimated as described for the validation data. Observed community compositions resembled those identified by prior 16S rRNA gene amplicon sequencing surveys but with substantial added complexity due to the enhanced taxonomic resolution provided by VIRGO2 ( Fig. 4a ). Our inclusion of metagenomes from multiple locations around the world allowed us to evaluate if vaginal bacteria display biases in their prevalence across geographies. We found that most bacterial species were common in metagenomes from every sampled continent, indicating they have cosmopolitan distributions ( Fig. 4a ). This includes both L. crispatus and L. iners , as well as most species that frequent CST IV communities ( Fig. 4a ). However, some species do exhibit biases in their geographic distribution. The sister species L. gasseri and L. paragasseri were both more prevalent in metagenomes from North American and European women than other geographies ( Fig. 4a , Supplementary Table 4). Whereas L. portuensis was common in L. crispatus and L. iners dominated metagenomes from women in Asia but rare elsewhere ( Fig. 4a , Supplementary Table 4). Finally, Ca . L. vaginae was rare in CST IV communities from Asia but common to these communities from other continents ( Fig. 4a , Supplementary Table 4). These results indicate that while most vaginal bacterial species exhibit cosmopolitan distributions, there are several species that do display biases in their geographic distributions. Download figure Open in new tab Fig. 4 Description of the taxonomic composition of the vaginal microbiome and mycobiome, as estimated using VIRGO2. a, VIRGO2 estimated taxonomic composition of the 2,560 metagenomes used in its construction. The geographic origin and community state type of the metagenomes are displayed above the bars. b, Pairwise correlation coefficients in the observed relative abundances of Gardnerella spp. Coefficients were estimated using ANCOMBC, and only significant values are displayed (p<0.001). Colored points along the diagonal represent the average relative abundance of the species observed in vivo . The species are ordered by the phylogeny produced for Fig, 2b and that phylogeny is again displayed to the left of the panel. c, VIRGO2 estimated taxonomic composition of the mycobiome for the 339 (13.2%) metagenomes where fungi were detected. The sample’s geographic origin and the community state type of the corresponding microbiome are displayed to the left of the bars, along with a heatmap representing the log 10 transformed total fungal relative abundance. One major advancement provided by VIRGO2 is its ability to distinguish the now 15 recognized species of Gardnerella . Of the 15, G. vaginalis , G. swidsinskii, and G. piotti were the most common in the analyzed metagenomes ( Fig. 4a ). Unlike the vaginal lactobacilli, Gardnerella spp. are often identified together in the vaginal microbiome. The median number of Gardnerella species identified in a metagenome was four and the maximum was twelve. To determine if there was any structure to the subcommunities of Gardnerella spp., we estimated correlation coefficients between the relative abundances of the individual species. While most pairwise correlations between Gardnerella spp. were not significant, some significant positive correlations (n=8) and negative correlations (n=23) were identified ( Figure 4b , p<0.001, ANCOMBC). Many of the identified correlations involved members of the subclade containing G. swidsinskii , G. leopoldii , and G. sp00358545. These three species had positive correlations with each other and negative correlations with many other Gardnerella species ( Figure 4b ) and G. swidsinskii alone accounted for eight of the observed negative correlations (34.7%. Subsequent investigation revealed that most G. swidsinskii genomes (88% of isolate WGS, 79% of MAGs) harbor a unique biosynthetic gene pathway for the production of a lacticin 481-like bacteriocin (Additional File 1). This bacteriocin could drive the numerous observed negative correlations between G. swidsinskii and other Gardnerella spp. The vaginal mycobiome as characterized by VIRGO2 A key new feature of VIRGO2 is its ability to characterize the vaginal mycobiome. During the construction of the catalog, genomic data of several fungal species were identified in the un-binned contigs of the metagenomes. These data were further supplemented with isolate whole genome sequences of relevant species. A total of 27,545 fungal NRGs are included in VIRGO2 belonging to six species including: Candida albicans , Nakaseomyces glabratus (formerly C. glabrata ), Pichia kudriavzevii (formerly C. krusei ), C. tropicalis , C. dubliniensis and C. orthopsilosis . To characterize the vaginal mycobiome, we extracted the relative abundances of these fungal species from the VIRGO2 mapping results. Fungi uncommonly comprise a significant proportion of the overall microbial community—only 339 (13.2%) of the metagenomes had a total fungal relative abundance over 10 -4 ( Fig. 4c ). L. iners -dominated bacterial communities were over-represented (n=110, 38.6%) in the subset of fungal-containing metagenomes ( Fig. 4c , logistic regression, p adj <0.01). We next examined the taxonomic composition of the mycobiome. C. albicans was the dominant fungal species in most of these metagenomes (n=285, 84.1%, Fig. 4c ). A smaller subset of metagenomes had a mycobiome dominated by either N. glabratus (n=47, 13.9%) or P. kudriavzevii (n=3, 0.9%). Mycobiomes that were dominated by N. glabratus tended to represent a higher proportion of their overall metagenome than those that were dominated by C. albicans ( Fig. 4c , anova, p adj =0.0045). However, both C. albicans and N. glabratus were found to occasionally represent a more substantial proportion of the overall metagenome (max 29.2% and 21.2%, respectively). Identification and characterization of bacteriophage in metagenomic data using VIRGO2 Bacteriophages, both as active virions and integrated prophages, are thought to be a significant component of the microbiome, but their abundance and role in vaginal ecology remains poorly described. To enable VIRGO2’s ability to characterize these viruses, we identified and annotated bacteriophage sequences in the metagenomic data used in the construction of the database. A total of 3,414 bacteriophage contigs encoding 89,793 of the NRGs were identified and then annotated with a taxonomy, lifestyle (temperate versus virulent), and predicted host (Supplementary Table 5). The relative abundance of bacteriophages was estimated using the VIRGO2 mapping results for the 2,560 metagenomes. We found bacteriophages typically represent 5-10% of the vaginal metagenome, with L. crispatus dominated communities having a slightly higher proportion of bacteriophages than other CSTs ( Fig. 5a ). However, there were a few instances where the proportion of bacteriophages was much higher (max observed 43.7%). The lifestyle annotations were then used to identify the 63 metagenomes (2.4%) where the proportion of virulent bacteriophages was greater than that of temperate ( Fig. 5b ). Four of these metagenomes originated from the same person, spread evenly over a 10-week longitudinal sampling period. The peak bacteriophage abundance occurred at Timepoint 4, when the bacterial community experienced a concurrent decline in the relative abundance of L. gasseri ( Fig. 5c ). We were able to visualize putative parts of a once intact bacteriophage by transmission electron microscopy (TEM) of filtrate of stored samples collected the day prior to Timepoint 4 ( Fig. 5d ). The ∼50kbp genome of the predicted virulent bacteriophage driving these observations was also identified in the metagenome assembly from the four timepoints and annotated using Pharokka ( Fig. 5e ). Lysogeny related genes were not identified in the genome, consistent with the predicted virulent lifestyle of the bacteriophage, and several genes encoding tail-fiber proteins were identified, consistent with the TEM micrograph. These results demonstrate how VIRGO2 can be used to identify and characterize bacteriophage populations in the vaginal microbiome. Download figure Open in new tab Fig. 5 Description of the vaginal phageome as estimated by VIRGO2 and the identification and characterization of a probable lytic bacteriophage bloom. a, Relative abundance of bacteriophage genes delineated by the community state type of the corresponding microbiome. b, Ratio of the relative abundance of predicted virulent to temperate bacteriophage genes. Points are colored by the community state type of the microbiome and only samples where the relative abundance of virulent bacteriophage surpassed that of temperate bacteriophage are shown. Four of the five metagenomes from a longitudinal collection of samples from the same woman were identified on this plot and are circled by dotted lines and labeled by sequential timepoints separated by ∼2 weeks. c, Taxonomic composition of corresponding microbiome at those five timepoints. The taxa color scheme is consistent and can be found in Figs. 2 , 3 , 4 and 6 . d, tunneling electron microscope (TEM) images of bacteriophage-like particles identified in the 0.22µm filtrate of a sample collected from the woman immediately prior to Timepoint 4. TEM images were altered only to increase the size and color of the scale bar. Original images with associated metadata banner can be found in Additional File 1. e, Gene annotations of a predicted lytic bacteriophage identified in the assemblies from Timepoints 1,3,4 and 5. The landscape of carbohydrate acting enzyme expression differs depending on community composition A particular strength of VIRGO2 is its ability to analyze and integrate vaginal metagenome and metatranscriptomic data. To demonstrate this feature, we mapped the sequence reads of 344 previously published vaginal metatranscriptomes (see Supplementary Table 6 for accession numbers) to VIRGO2 and then extracted the expression of genes predicted to act on carbohydrates using the available CAZy annotations. The carbohydrate-active expression profiles were then embedded into 2D space using PacMAP 78 , revealing two distinct clusters ( Fig. 6a ). The first cluster was primarily comprised of microbiomes assigned to CST IV-A, IV-B and II, while the second was comprised of microbiomes assigned to CSTs I, III, and V. We next performed a differential gene expression analysis to identify the specific carbohydrate activities driving the separation of these two clusters. Many of the identified differentially expressed (DE) CAZy functions were predicted to be involved in the catabolism of either glycogen or mucin-associated glycans ( Fig. 6b ). For example, expression of GHs 20 (β-N-acetylglucosaminidases), 29 (fucosidase), 33 (sialidase), and 101 (endo-α-N-acetylgalactosaminidase) were all higher in samples from cluster 1 whereas GHs 4 (β-glucosidase), 13 (pullulanase), 26 (mannanase), 31 (α-galactosidase), 65 (maltose phosphorylase) were more highly expressed in samples from cluster two ( Fig. 6b ). The integrated taxonomic and functional annotations of VIRGO2 enabled us to identify the species responsible for the expression of these enzymes. Many of the defining characteristics of cluster 1 were largely expressed by various species of Prevotella and Sneathia ( Fig. 6c ), whereas the lactobacilli were mostly responsible the expression of cluster 2 associated enzymes. This result shows that vaginal microbial communities likely metabolize host-produced glycogen but access these resources using different enzymes. However, catabolism of mucin glycans appears to be an activity better accomplished by the microbes that frequent CST IV, due to their expression of enzymes that remove terminal fucose and sialic acid residues. Download figure Open in new tab Fig. 6 Description of the identification of differences in the expression of carbohydrate-active enzymes by the vaginal microbiome. a, Two-dimensional PacMap embedding of the in vivo expression patterns of carbohydrate-active enzymes extracted from the VIRGO2 mapping results of 363 vaginal metatranscriptomes. Points are colored by the community state type of the sample. Two clusters were observed. b, Volcano plot displaying the results of a differential gene expression analysis comparing clusters 1 and 2. Enzymes are colored by their predicted substrate with a particular emphasis placed on mucin and glycogen. Markers designate the class of the enzyme (hydrolase, transferase, or other). c, Taxa driving the expression of significantly differentially expressed carbohydrate-active enzymes in clusters 1 and 2. The displayed bars represent the average contribution of each taxon to the expression of the enzyme in samples from the corresponding cluster. Berryella sp001552935 is a common yet under-characterized member of the vaginal microbiome In our initial MAG survey, we found Berryella sp001552935 to be an unexpectedly common member of the vaginal microbiome. Subsequent analysis demonstrated that this species had a near universal presence in CSTs IV-A (98.3%) and IV-B (93.0%) communities but rarely comprised more than 10% of these vaginal microbiomes ( Fig. 7a ). The relative abundance of Berryella sp001552935 was also highly diagnostic of CST IV-A and IV-B (receiver-operating characteristic curve, AUC=0.94, Supplementary Fig. 4). To better understand the biology of Berryella sp001552935 in the vaginal microbiome, we examined the functions encoded in its genome. The species lacks predicted phosphotransferase systems and appears to be largely asaccharolytic. Acetate and lactate are predicted to enter the cell through a FocA-like porin and then serve as precursors for biosynthetic pathways ( Fig. 7b ). Energy production, on the other hand, appears to be primarily driven by the import and breakdown of formate and arginine ( Fig. 7b ). Catabolism of arginine is also predicted to result in the production of the biogenic amine putrescine. The in vivo expression of the enzymes in these pathways were confirmed and quantified by extracting the transcriptome of Berryella sp001552935 from VIRGO2 mapping results for five metatranscriptomes where the species was well represented ( Fig. 7c , minimum 250,000 Berryella sp001552935 reads). All enzymatic steps were expressed in vivo with average values ranging from 10 2 to 10 4 transcripts per million ( Fig. 7d ). These results give insight into the ecological role of Berryella sp001552935 in the vagina and highlight how VIRGO2 can be used to extract and analyze a species’ transcriptome from metatranscriptomic data. Download figure Open in new tab Fig. 7 Characterization of the abundance and metabolism of Berryella sp001552935 in the vagina. a, Log 10 transformed relative abundance of Berryella sp001552935 delineated by the community state type of the sample. b, Diagram of predicted metabolic pathways encoded in the genome of Berryella sp001552935. Enzyme number designations correspond to: (1) acetate kinase, (2) phosphate acetyltransferase, (3) pyruvate phosphate dikinase, (4) L-lactate dehydrogenase, (5) NAD-dependent malic enzyme (6) hydrolyase, tartrate alpha subunit/fumarate domain protein, Fe-S type, (7) argininosuccinate lyase, (8) argininosuccinate synthase, (9) ornithine carbamoyltransferase, (10) carbamate kinase, (11) diaminopimelate decarboxylase, (12) formate dehydrogenase H. c, Relative abundance of Berryella sp001552935 in the five metagenomes for which a corresponding metatranscriptome had >250,000 reads mapping in VIRGO2 to the species. The five samples originate from five different North American women and are assigned separate colors, which appear underneath the bars. Berryella sp001552935 is highlighted by a dotted outline and all taxa are colored according to the legend found in Figs 2 , 3 , 4 and 6 . d, In vivo expression of the enzymes diagramed in subpanel b extracted from the VIRGO2 mapping results for the five metatranscriptomes containing Berryella sp001552935. Black bars represent the average expression, and the five points are colored according to the samples characterized in panel c. Discussion Genomic databases serve as a critical resource for the robust and reproducible analysis of microbial ‘omics data. Here, we present the construction, validation, and application of VIRGO2, an enhanced iteration of our non-redundant gene catalog for the vaginal microbiome. VIRGO2 was constructed ab initio from a set of 2,560 shotgun metagenomes and 3,176 isolate genome sequences and is a complete overhaul of the original database. Sequence data from five continents were used to build VIRGO2, providing a substantial improvement in the database’s geographic representation. VIRGO2 includes twice as many non-redundant genes as VIRGO 20 and catalogs the genomic diversity of most common and uncommon vaginal bacteria, fungi, and viruses. Over 90% of the genes in VIRGO2 have taxonomic annotations and 80% have at least one functional annotation. The inclusion of data from both MAGs and the residual un-binned contigs also ensures that sequence data from unknown or unidentified organisms are included in the catalog and provides the opportunity for their future annotation. VIRGO2 is also not just a genomic database and its associated annotations. Software to analyze shotgun metagenomes and metatranscriptomes is bundled with the database and was validated to accurately characterize the vaginal microbiome. We anticipate that VIRGO2, and its subsequent releases, will continue to provide a critical resource and tool for the analysis of vaginal microbiome omics data. A major limitation of the initial VIRGO1 database was the lack of species-level taxonomic resolution within the genus Gardnerella . Concurrent with the construction and publication of VIRGO, G. vaginalis was split into thirteen species, of which six were given names 64 , 79 . This reclassification has proven important because diversity within Gardnerella has been associated with symptomatic BV 47 and with preterm birth 27 , 48 , 80 . In this iteration of VIRGO, we cataloged the genomic diversity of all described Gardnerella species and identified two additional species. Our subsequent analysis reaffirmed that most women with Gardnerella in their VMB have several species represented 47 , 48 , 80 , 81 , and demonstrated that the subcommunity of Gardnerella spp. has structure 20 , 47 . This is somewhat unexpected given that these closely related species likely have substantial overlap in their niche use 82 . The vaginal Lactobacillus species, which bear a similar degree of relatedness to one another, do not partition their niche space to the same degree as species of Gardnerella 33 . It remains to be determined if Gardnerella species exhibit mutualistic interactions that drive their co-occurrence 83 , if they engage in resource or spatial niche partitioning 84 , or if they simply cannot exclude one another from the vagina 85 . The limited number of host-produced nutrients (e.g. glycogen, mucin) available to the vaginal microbiome, may preclude resource partitioning as a major driver of Gardnerella spp. coexistence. We did find examples of negative correlations between the abundance of Gardnerella species, many of which involved G. swidsinskii . The genome of this species was found to harbor a unique biosynthetic pathway for the production of a lantibiotic that may be responsible for these negative associations. Future studies, including in vitro experimentation, should endeavor to characterize interactions between Gardnerella species and how these important bacteria relate to vaginal dysbiosis. Topic models, as employed previously 86 , would naturally lend themselves to the investigation of Gardnerella subcommunity structure. We identified several novel bacterial species in the collection of MAGs used to build VIRGO2. Many of these species were close relatives of known, well described, vaginal bacteria (e.g. novel species in the genera Gardnerella ) and their biology likely resembles that of their better described sister species, with perhaps some unique aspects. However, Berryella sp001552935 stood out as a near-universal member of the CST IV microbiota that is not often discussed in the literature. The species is in the family Eggerthellaceae and may have been referred to as “ Eggerthella sp.” or “ Eggerthella type 1” in a few 16S rRNA gene sequencing-based studies 25 , 87 – 91 . Twice this “Eggerthella type 1” species has been associated with all four clinical signs of the Amsel criteria for diagnosis of bacterial vaginosis 88 , 89 , concordant with our finding of Berryella sp001552935’s diagnostic value for CST IV. Due to the species’ specificity for non- Lactobacillus dominant microbiota, it is difficult to disentangle its association with vaginal health from that of other BV- associated microbes. Nevertheless, one study did find “ Eggerthella type 1” to be correlated with higher IL-1⍺, IL-1β, and TNF-⍺, and two found it to be associated with increased risk of HIV acquisition 25 , 91 . Based on its genomic contents and in vivo transcriptional activity, we predict Berryella sp001552935 to be largely asaccharolytic, like its relative E. lenta 92 , and to similarly rely on acetate and arginine as sources of carbon and energy. Acetate is the end-product of the fermentative activities of several CST IV community members 93 and is one of the most common molecules in the vagina when lactobacilli are not present 39 . Arginine is also a common component of vaginal fluid and can be found at high concentrations in seminal fluid 94 . The catabolism of arginine by Berryella sp001552935 is predicted to result in the production of the biogenic amine putrescine, which can function to raise vaginal pH and likely contributes to vaginal malodor. We therefore posit that Berryella sp001552935 acts as a decomposer in the vaginal ecosystem, upcycling waste products into biomass and energy and aiding the growth of other BV-associated microbes. We placed an emphasis on cataloging the non-bacterial components of the vaginal microbiome in VIRGO2 to expand the utility of the database. Several fungal species were identified in the shotgun metagenomics data used to build VIRGO2 and more were added through the inclusion of isolate whole genome sequences. Our analysis of the shotgun metagenomes revealed that C. albicans , N. glabratus and P. kudriavzevii were the primary components of the vaginal mycobiome, concordant with past survey data 95 – 97 . Some studies have identified other species as minor components of the fungal communities 98 , 99 ; however, we found no evidence of these species in the shotgun metagenomes. It could be that they are present but only below the limit of detection of metagenomics (<0.01% relative abundance), or they could simply represent fungal contaminants from the air or molecular reagents 100 . We also found that only 13.2% contained a detectable abundance of fungi, a rate similar to past microscopy-derived estimates of asymptomatic Candida carriage 101 , 102 . These fungi containing metagenomes were more likely to have a bacterial community that was dominated by L. iners, than any other VMB compositions. It is likely that detection of fungi in metagenomic data is harder in CST IV bacterial communities due to their higher population sizes 103 . Past observational studies have also found associations between L. crispatus 104 , 105 or L. iners 43 and detection of vaginal yeast. Symptomatic vulvovaginal candidiasis (VVC), on the other hand, has been repeatedly linked to BV and BV indicated antibiotic therapy 106 . It is commonly thought that lactobacilli temper the virulence of vaginal fungi, although the exact mechanism has yet to be elucidated. The inclusion of fungal sequence data in VIRGO2 enables the characterization of the vaginal mycobiome using metagenomic and metatranscriptomic data and will advance our understanding of its relationship to the bacterial microbiome. Our efforts to include, identify, and catalog viral sequences in VIRGO2 enable its use for the characterizing the vaginal virome. We found that bacteriophages were responsible for the majority of the virome, with eukaryotic viruses (e.g. human papillomavirus) representing only a minor fraction, consistent with past studies 107 , 108 . Bacteriophage have long been theorized to be major drivers of compositional changes in vaginal bacterial communities 109 . Many women have Lactobacillus -dominant bacterial communities; a virulent bacteriophage capable of infecting the dominant Lactobacillus could collapse its population, resulting in a dramatic shift in community composition. Yet, we found that most bacteriophage in the vagina are predicted to have temperate lifestyles. A high abundance of virulent bacteriophages was observed in only a handful of metagenomes, and the bacterial communities in these cases tended not to be dominated by lactobacilli. These results suggest that virulent bacteriophage are not major drivers of bacterial community composition and are likely not responsible for most transitions away from Lactobacillus dominance. Temperate bacteriophage integrate into their host bacterium’s genome and excise themselves in response to a variety of environmental or metabolic triggers 110 . Vaginal Lactobacillus are known to harbor numerous prophage in their genomes 111 – 113 , but the conditions that induce their excision are not well characterized 114 . Induction of temperate bacteriophage could impact community composition by reducing the population size of their host 115 ; however, prophage can also function to protect their host from infection by other bacteriophage 116 , regulate host gene expression 117 , kill closely related competitors 118 , regulate population density 119 , or act as agents of horizontal gene transfer 120 . Going forward, VIRGO2 will facilitate the characterization of the temperate bacteriophage common to the VMB and how they influence community dynamics and vaginal health. To demonstrate how VIRGO2 can be used to analyzed vaginal metatranscriptomes, we examined the expression of carbohydrate acting enzymes in a previously published dataset 121 . The results from this analysis identified several CAZy families that were differentially expressed between communities that were dominated by Lactobacillus versus those that were not. Many of these differentially expressed enzyme families were predicted to be involved in the breakdown of either glycogen or mucin-associated glycans. This result suggests that, while vaginal bacteria have substantial overlap in their resource use, the enzymatic functions they use to access these resources, and the extent to which they utilize them, are different. For example, communities containing Prevotella and Gardnerella spp. might be more capable of catabolizing mucin glycan chains because they express the enzymes necessary to remove terminal sialic acid and fucose residues, providing access to the mucin-associated glycans 50 , 122 . The vaginal lactobacilli do not express these enzymes and therefore may only have limited access to those mucin glycan chains. Glycogen, on the other hand, appears to be a major source of carbon and energy for most vaginal bacteria but is catabolized using slightly different enzymatic functions. Lactobacilli primarily use GH13 and GH33 domain containing enzymes, whereas many of the bacteria that comprise CST IV communities (e.g. Gardnerella spp., Prevotella spp., or Sneathia spp.) express GH77 and GT35 domain containing enzymes. These differences could be harnessed to selectively promote the growth of lactobacilli or inhibit the growth of non-optimal bacteria. For example, a recent study by Pelayo et al. identified several small molecule inhibitors of Prevotella and Gardnerella sialidases that could limit the growth of these bacteria in the vagina 122 . Another study found that supplementation with maltose, a metabolite of glycogen produced by vaginal lactobacilli 123 , was able to shift the macaque vaginal microbiota to Lactobacillus- dominance 124 . Additional studies aimed at understanding microbial resource use in the vagina, including beyond carbohydrate metabolism characterized here, could reveal additional means to modulate community composition. Integrative multi-omic analyses can provide mechanistic insight into the interactions between humans and their resident microbes. Yet a major barrier to this discovery is the lack of computational resources for the analysis of these data. VIRGO2 helps fill this critical need by providing an analytical framework for the analysis and integration of vaginal metagenomic and metatranscriptomic data. Here, we present a major update to this resource that 1) expands the geographic coverage of the constituent sequence data, 2) broadens the scope of the database to include non-bacterial microbes including fungi and viruses, and 3) remasters the taxonomic and functional annotations to match current standards. Still, VIRGO2 is not without limitations. Although sequence data from five continents were included in its construction, most continents are only represented by a single dataset from a single country, and South American populations are not represented. The database is also missing representation of pre- and post-reproductive age individuals as no shotgun metagenomes were available for these populations. Despite these limitations, we expect VIRGO2 to enhance the analysis of vaginal microbiome ‘omics data to better understand how these not-so-simple microbial communities impact health. Author Contributions M. T. France, Conceptualization, Methodology, Software, Validation, Data curation, Formal analysis, Writing – original draft, Writing – review and editing | I. Chaudry, Software, Formal analysis, Validation, Writing – review and editing | L. Rutt, Investigation, Data curation, Supervision, Writing – review and editing | M. Quain, Investigation, Writing – review and editing | B. Shirtliff, Investigation, Writing – review and editing | E. McComb, Investigation, Writing – review and editing |F. A. Hussain, Formal analysis, Writing – review and editing | A. Maros, Formal analysis, Writing – review and editing | M. Alizadeh, Formal analysis, Writing – review and editing | MA. Elovitz, Funding Acquisition, Supervision, Writing – review and editing | D. Relman, Funding Acquisition, Supervision, Writing – review and editing | A. Rahman, Funding Acquisition, Supervision, Writing – review and editing | R. M. Brotman, Funding Acquisition, Supervision, Writing – review and editing | J. Price, Funding Acquisition, Supervision, Writing – review and editing | M. Kassaro, Funding Acquisition, Supervision, Writing – review and editing | J. Holm, Formal Analysis, Supervision, Writing – review and editing, | B. Ma, Conceptualization, Writing – review and editing | Jacques Ravel, Conceptualization, Funding acquisition, Supervision, Writing – review and editing. All authors contributed and approved the final version of the paper. Declaration of interests J.R. is a cofounder of LUCA Biologics, a biotechnology company focusing on translating microbiome research into live biotherapeutic drugs for women’s health. M. A. E. is a consultant and has equity in MIRVIE, a company focused on cell free RNA and adverse pregnancy outcomes. R. M. B.’s research is supported by in-kind donation from Hologic. All other authors declare no competing interests. Resource availability Lead contact Additional information and requests for additional data and code should be directed to and will be fulfilled by the Lead Contact, Jacques Ravel ( jravel{at}som.umaryland.edu ) Data and code availability The VIRGO2 software, the non-redundant nucleotide gene catalog and database, the VOG amino acid protein family database, the associated annotation files, and detailed tutorials are available at: http://virgo.igs.umaryland.edu . The metagenomes used to construct VIRGO2 are all publicly available can be obtained from either the National Center for Biotechnology Information (NCBI) Sequence Read Archive (SRA) or from the European Nucleotide Archive using the accession numbers provided in Supplementary Table 1. All scripts used in the described analyses and to construct the displayed figures are available at: https://github.com/ravel-lab/VIRGO2 . Methods Generation of metagenome assembled genomes A total of 2,560 shotgun metagenomes were used in the generation of the VIRGO2 database (Supplementary Table 1). The vast majority of these metagenomes were generated from human vaginal swab or lavage specimens (n=2503, ∼98%). However, a small proportion were generated from male urethral swab specimens, which were generated as part of a sexual network study (n=57, ∼2%). Sequence reads originating from the human genome were identified and removed in all shotgun metagenomes using BMTagger ( ftp://ftp.ncbi.nlm.nih.gov/pub/agarwala/bmtagger , default settings). This included metagenomic datasets downloaded from public databases to ensure all host reads were removed prior to processing. The host-removed datasets were further processed using fastp 125 to remove barcode sequences and trim low-quality bases (v0.21.0, -l 50 - W 4 -M 20). The median number of reads per metagenome post-processing was 1.27×10 7 (range from 2.0×10 4 to 3.3×10 8 ) and only 3.4% of metagenomes had fewer than 10 6 reads. Metagenomes originating from studies performed in Fiji, Ireland, and China were not sequenced to the same depth as the rest of the metagenomes and therefore had fewer reads per sample on average (median: 1.6×10 6 ; range: 2.0×10 4 -4.8×10 7 ). The metagenomes sequenced from male urethral swabs also had a greater proportion of host-reads resulting in fewer available microbial reads (median: 1.4×10 6 , range: 2.8×10 5 -6.3×10 7 ). De novo assembly was attempted on all shotgun metagenomes using metaspades 126 (v3.15.3, default settings) and 2,548 (99.5%) produced assemblies. The raw assemblies were then separated into a metagenome assembled genomes (MAGs) using a custom multipronged approach whereby binning was performed on each metagenome using MaxBin2 127 (v2.2.7, -min_contig_length 500), metaBat2 128 (v2.12.1, -- maxEdges 500), and Metadecoder 129 (v1.0.8, default settings). The resulting set of bins were then compiled using metaWRAP 130 , which relies on checkM 131 to estimate MAG completeness and degree of contamination. All MAGs were further processed using MAGpurify 132 (v2.1.2, default workflow) and cleaned versions were accepted where the degree of contamination, as estimated using checkM, decreased, without a concomitant decline in estimated completeness >1%. MAGs demonstrating a final completion level >70% and <10% contamination were used in the analysis and in the construction of the VIRGO2 database (n=15,142, Supplementary Table 2). A total of 21,202,535 genes were then identified on the MAGs using Prodigal 133 (v2.6.3, -c -p single). Processing residual metagenome assembly contigs Contigs greater than 2,000 bp that were not incorporated into any bin, were processed as “un-binned contigs”. These contigs likely represent either fragments of bacterial genomes which were not successfully binned or fragments of non-bacterial genomes (e.g. fungi, bacteriophage). A total of 8,294,162 genes were identified on the un-binned contigs using Prodigal 133 (v2.6.3; -c -p meta). Cultured isolate genome (CIG) sequences Whole genome sequences of known vaginal bacteria, fungi, and viruses were used to supplement the metagenomic data in the construction of the VIRGO2 catalog. A total of 4,013 genomes were downloaded representing 198 species, of which 190 were bacteria, 4 were fungi ( Candida albicans , Candida dubliniensis , Nakaseomyces glabratus , and Pichia kudriavzevii ), 3 were viruses (human immunodeficiency virus, human papilloma virus, and human alpha herpes virus), and 1 was a protist ( Trichomonas vaginalis ). A complete list of the species targeted, and accession numbers for these genome sequences can be found in Supplementary Table 3. Sequences were acquired from NCBI in February of 2023. A total of 8,015,873 genes were identified on the CIGs using Prodigal 133 (v2.6.3, -c -p single). Generation of non-redundant gene clusters All genes identified in the three data sources (MAGs: 21,202,535, un-binned contigs: 8,294,162, CIGs: 8,015,873) were clustered using CDHIT 134 (v4.8.1, cd-hit-est -n 8 -c 0.95 -aS 0.9) with thresholds of 95% sequence identity, and 90% overlap of the shorter sequence. This process produced 1,880,579 gene clusters, each represented by their longest sequence constituent and hereafter referred to as “non-redundant genes”. The median number of sequences per cluster was 2 (range 1-6,464), and 729,189 clusters (38.8%) were comprised of single sequences. Identifying and removing anomalous non-redundant genes Non-redundant genes were searched against the human genome 135 (NC_060925.1, T2T- CHM13v2.0) and against the SARS-CoV-2 genome 136 (NC_045512.2) using blastn 137 (v2.10.0, - evalue 0.00001) to identify (percentage identity ≥75%, coverage ≥ 50%) and remove these undesired sequences from the VIRGO2 database. A total of 34,321 non-redundant genes were removed due to matches to the human genome and 20 due to matches to the SARS-CoV-2 genome. An additional 9,159 non-redundant genes which contained at least one “N” in their sequence were removed. An all-versus-all blastn (v2.10.0, -evalue 0.00001) search was then performed against the remaining non-redundant genes to identify those which contain a stretch of sequence, of at least 100bp, identical to that found in another non-redundant gene. For these cases of exact contiguous matches, only the longer gene was kept, resulting in the removal of an additional 37,665 genes. A final filtering step was applied to identify non-redundant genes whose prevalence, abundance, and tetramer content were dissimilar from that of the overall species’ distribution. These genes were primarily repeats of adapter or barcode sequence and were removed from the database (n=26,897). The final number of VIRGO2 non-redundant genes following these filtering steps was 1,772,517. Assigning taxonomy to VIRGO2 non-redundant genes The taxonomy of the VIRGO2 source sequences (MAGs, CIGs, un-binned contigs) form the basis of the annotations applied to the non-redundant gene clusters. Taxonomy was first assigned to the MAGs and CIGs using GTDB-tk 138 (v2.1.1, GTDB release 207_v2, classify_wf). Then, a two-stage strategy was used to assign taxonomy to the un-binned contigs. First, the un-binned contigs were searched against the NCBI nt database (accessed March 2023) using blastn (v2.10.0, -evalue 0.00001). Second, the un-binned contigs were searched, using the same settings, against a blast database constructed from the MAGs and CIGs. Taxonomic annotations were propagated to the unbinned contigs preferentially from the MAGs and CIGs over those from matches to the NCBI nt database. The taxonomic annotations from the source sequences were then applied to the final non-redundant gene clusters based on the taxonomy of their constituents using the following rules: (1) gene clusters with uniform species-level taxonomy were assigned to that singular species (n=1,193,357); (2) gene clusters with variable species-level taxonomy, but a majority (>50%) of one, were assigned to that species (n=279,714); (3) gene clusters without a majority species-level taxonomy but a majority (>50%) genus-level taxonomy, were assigned to that genus (n=71,195); (4) genes without a majority genus-level taxonomy were assigned as “MultiGenera” genes (n=68,548); (5) genes without any available taxonomic information were not assigned any taxonomic annotation (n=159,703). Assigning function to VIRGO2 non-redundant genes A rich suite of functional annotation schemes was applied to the non-redundant genes. EggNOG- mapper 65 (v2.1.10, eggNOG database v5.0.2, -m diamond, -evalue=0.001) was used to annotate the genes with the following schemes: eggNog 68 , COG 139 , PFAM 70 , GO terms 71 , 140 , EC numbers 141 , BiGG Reaction 73 , KEGG 74 , and Carbohydrate-active enzymes 75 (CAZy). Antibiotic resistance genes and virulence factors were identified and annotated using AMRFinderPlus 67 (v3.10.16, database v2022-11-2, default settings). Secondary metabolite biosynthetic gene clusters were identified in the MAGs using antiSmash (v7.1.0, --genefinding-tool prodigal, --clusterhmmer, -cc-mibig, --cb- general, --cb-subclusters, --cb-knownclusters) and then the annotations were propagated to the appropriate NGR cluster 76 . The VIRGO2 non-redundant genes were translated into protein sequences using biopython (v1.81) and the appropriate codon table. Orthologous protein sequences were identified for each species, individually, with OrthoFinder 142 (v2.5.4, default settings). The 1,772,517 non-redundant genes were reduced to 1,032,117 orthologs. Identifying and annotating bacteriophage sequences in VIRGO2 Bacteriophage sequences among the raw metagenome assemblies were identified using virSorter2 143 (v2.2.4, --min-length 3000 --min-score 0.5) and their quality was assessed using checkV 144 (v1.01, database v1.5, default settings). Predicted bacteriophage sequences were only used when they met the following conditions: (1) more than one known viral gene detected, (2) sequence not predicted to be a prophage, (3) fewer than 10 bacterial host genes detected, and (4) checkV completeness estimate > 90%. This conservative filtering process produced 3,414 predicted bacteriophage genome sequences which encoded 43,365 of the VIRGO2 non-redundant genes. Bacteriophage lifestyles (lysogenic versus virulent) were predicted using PhaTYP 145 (v1, default settings) and the sequences were classified at the family level using PhaGCN 146 (v1, default settings). Bacteriophage host identity was predicted using Prokaryotic virus Host Predictor 147 (PHP, v1, default settings) with a dereplicated set of the VIRGO2 MAGs (dereplication performed using dRep 148 , v3.1.1, --sa 0.975, --S_algorithm fastANI, --clusterAlg ward) used as the host database. PHP host predictions were accepted when the reported score was ≥ 1400 and were recorded at the genus level. Bacteriophage taxonomy, host, and lifestyle annotations were propagated to the appropriate non-redundant genes (Supplementary Table 5). Using VIRGO2 to profile In silico generated mock communities, vaginal shotgun metagenomes and, metatranscriptomes Software to map reads to the VIRGO2 non-redundant gene catalog and to characterize metagenome/metatranscriptome composition is provided along with the catalog and is available online (virgo.igs.umaryland.edu). Briefly, reads are mapped to the non-redundant gene catalog using Bowtie2 (v4.1.2, -N 0,-L 20 -D 20,-R 3,-k 10,--sam-no-qname-trunc,--end-to-end,--seed 343, --no-unal) and the resulting sam file is parsed using samtools (v1.11, sort, view commands) and custom python3 code. To validate VIRGO2’s ability to characterize vaginal communities, we generated mock shotgun metagenomes using inSilicoSeq (v1.5.3, --seed 343, -n 10m, --model novaseq) 149 . Isolate genome sequences of several vaginal bacteria were used to generate mock shotgun metagenomes comprising known abundances of species. The exact isolates used are denoted in Supplementary Table 3. Accuracy of the resulting quantifications was assessed using a linear regression of observed versus expected relative abundances of bacteria using the scipy python package (v1.1.10). Following these validation efforts, the 2,560 metagenomes used to construct VIRGO2 and as well as 363 vaginal metatranscriptomes (Supplementary Table 1) were mapped to the catalog using the same approach. The results of these mappings were used in the subsequent analyses generating Figures 4 - 7 (metagenomes: Figures 4 , 5 , and 7 ; metatranscriptomes: Figures 6 and 7 ). Relative abundances were quantified using the sum of gene coverages for each bacterium, fungi, or virus. Identifying correlations in the abundance of Gardnerella spp The number of Gardnerella species identified in each metagenome was determined using a conservative 1% relative abundance threshold for detection. Correlations in the relative abundance of Gardnerella species were quantified using ANCOMBC (v2.0.3, secom_linear, method=pearson, thresh_hard=0.3) 150 , 151 . A lantibiotic operon was identified in the genome of G. swidsinskii C0102A2 using BAGEL4 (v.1.2, default settings) 152 . The lantibiotic gene was then identified in the genomes and MAGs of other G. swidsinskii using blast searches (blastp, v2.13.0, default settings) 137 . Characterization of vaginal bacteriophage Genes annotated as originating from bacteriophage were identified using the appropriate annotation table included in VIRGO2 (10.VIRGO2.phage.txt, available at: virgo.igs.umaryland.edu). The overall relative abundance of bacteriophage was quantified using the sum of gene coverages for the bacteriophage genes divided by the sum of coverages for all genes. To quantify the ratio of virulent to temperate bacteriophage, the sum of coverages of the bacteriophage genes annotated as virulent was divided by the sum of coverage of those annotated as temperate. Four of the 63 metagenomes with higher coverage of virulent than temperate genes were identified as originating from the longitudinal sampling of a single person. The high coverage of virulent bacteriophage genes was traced back to a single high coverage bacteriophage contig. The contig’s coding sequences were identified and annotated with Pharokka (v1.7.3, -g prodigal-gv, --dnaapler) 153 . Bacteriophage was isolated from a vaginal swab specimen collected one day prior to fourth timepoint, that had been stored at −80℃ without any preservation media 53 . The sample was eluted off the swab by vortex in 2 mL of sterile phosphate-buffered saline (PBS) for 2 min. The elution was sequentially filtered through a 5 µm, 0.45 µm, and 0.22 µm diameter filters (Thermo Fisher Scientific, Waltham, MA, USA). The final filtrate was used to prepare a negative stain TEM grid at the University of Maryland School of Medicine (UMSOM) Electron Microscopy Core. A formvar/carbon coated TEM grid underwent plasma cleaning (glow discharge) before 5 µL of the eluted bacteriophage sample was applied to formvar side of the grid. The sample was given 1 min to adsorb to the grid, and excess sample was removed. Next, a fixative (4% glutaraldehyde in 0.12 M sodium cacodylate buffer, pH = 7.2) was added to the grid for 5 min, and then washed four times with deionized water. Lastly, the grid was stained with 1% aqueous uranyl acetate for 1 min. Grids were dried at room temperature overnight before mounted on the TEM. Micrographs were collected on the FEI tecnai T12 instrument (Field Electron and Ion Company, Hillsboro, OR, USA). Analysis of carbohydrate-active enzyme expression The transcriptional activity of the analyzed 344 vaginal metatranscriptomes was quantified as transcripts per million (TPM). The expression of genes encoding carbohydrate-active enzymes 75 were identified and extracted using the VIRGO2 CAZy annotation table (8.VIRGO2.CAZy.txt, virgo.igs.umaryland.edu). The resulting dataset of CAZy enzyme expression was processed using the Pairwise Controlled Manifold Approximation (PacMap) dimensionality reduction method (v0.7.0, n_components=2, MN_ratio=0.5, FP_ratio=2.0, distance=angular), revealing two clusters. Differential expression analysis was conducted using VariancePartition (v1.28.9, default settings) 154 between the two groups to identify the CAZy enzyme classes responsible for the clustering. Characterization of Berryella sp001552935 The relative abundance of Berryella sp001552935 in samples from different CSTs was quantified as described above. Metabolic pathways were outlined via the KEGG orthologs assigned to a representative Berryella sp001552935 MAG (MAG14846, Supplementary Table 2) by BlastKOALA (v3.1, prokaryotes) 155 and then matched to VIRGO2 genes via blastp (blastp, v2.13.0, default settings) 137 . Five of the 363 vaginal metatranscriptomes were identified as having more than 100,000 reads mapping to Berryella sp001552935 genes. These samples were used to estimate the expression of genes in the putative metabolic pathways used by the bacterium to generate energy and assimilate carbon in the vagina. Supplementary Table Legends Supplementary Table 1: Accession numbers of metagenomes included in VIRGO2 Supplementary Table 2: Inventory of metagenome assembled genomes generated in this study Supplementary Table 3: Accession numbers of isolate whole genome sequences included in VIRGO2 Supplementary Table 4: Prevalence of vaginal bacteria across different geographies Supplementary Table 5: Inventory of bacteriophage genome sequences identified in VIRGO2 Supplementary Table 6: Accession numbers of vaginal metatranscriptomes used in our analyses Supplementary Figure Legends Supplementary Figure S1: Histogram displaying the number of non-redundant genes by the size of the cluster Supplementary Figure S2: Percent of VIRGO2 non-redundant genes assigned to various functional annotation schemes Supplementary Figure S3: Metagenome accumulation curves of prominent vaginal bacteria assessed at the level of orthologs (VOGs) Supplementary Figure S4: Receiver operating characteristic curves demonstrating the ability of the relative abundances of select microbial species ( Berryella sp001552935, G. vaginalis , G. swidsinskii , P. timonensis , F. vaginae , M. lornae , and A. indicium ) to predict an overall community composition assigned to CST IV-A or IV-B. Additional File 1: Fasta file containing the coding sequences of the Lantibiotic operon identified in G. swidsinskii Additional File 2: Original, unaltered TEM micrographs used in Figure 5 . Acknowledgements Research reported in this publication was supported in part the National Institute of Allergy and Infectious Diseases (NIAID), the National Institute for Nursing Research (NINR) of the National Institutes of Health (NIH) under awards UH2AI083264, R01AI116799, U19AI084044, R01NR014784 and R01NR015495, and from the Gates Foundation under awards OPP1189217, INV048956, and INV-048982. I. C. was supported by NIAID training grant T32AI162579. D.A.R. was supported by the Thomas C. and Joan M. Merigan Endowment at Stanford University. We thank the UMSOM Electron Microscopy Core Imagining Facility for assistance with electron microscopy. We also thank members of the Vaginal Microbiome Research Consortium (VMRC) for helpful discussions. Footnotes https://github.com/ravel-lab/VIRGO2 References ↵ Ruff , W. E. , Greiling , T. M. & Kriegel , M. A . Host-microbiota interactions in immune-mediated diseases . Nat Rev Microbiol 18 , 521 – 538 ( 2020 ). doi: 10.1038/s41579-020-0367-2 OpenUrl CrossRef PubMed ↵ Arnolds , K. L. & Lozupone , C. A . Striking a Balance with Help from our Little Friends – How the Gut Microbiota Contributes to Immune Homeostasis . Yale Journal of Biology and Medicine 89 ( 2016 ). ↵ Blacher , E. , Levy , M. , Tatirovsky , E. & Elinav , E . Microbiome-Modulated Metabolites at the Interface of Host Immunity . J Immunol 198 , 572 – 580 ( 2017 ). doi: 10.4049/jimmunol.1601247 OpenUrl Abstract / FREE Full Text Rastelli , M. , Cani , P. D. & Knauf , C . The Gut Microbiome Influences Host Endocrine Functions . Endocr Rev 40 , 1271 – 1284 ( 2019 ). doi: 10.1210/er.2018-00280 OpenUrl CrossRef ↵ Glick , V. et al. Vaginal lactobacilli produce anti-inflammatory β-carboline compounds . Cell Host Microbe ( 2024 ). doi: 10.1016/j.chom.2024.09.014 OpenUrl CrossRef ↵ Hwang , I. Y. & Chang , M. W . Engineering commensal bacteria to rewire host-microbiome interactions . Curr Opin Biotechnol 62 , 116 – 122 ( 2020 ). doi: 10.1016/j.copbio.2019.09.007 OpenUrl CrossRef Bai , X. et al. Engineering the gut microbiome . Nature Reviews Bioengineering 1 , 665 – 679 ( 2023 ). doi: 10.1038/s44222-023-00072-2 OpenUrl CrossRef Lamont , R. J. , Hajishengalllis , G. & Koo , H . Social networking at the microbiome-host interface . Infection and Immunity 91 , e00124 – 00123 ( 2023 ). OpenUrl ↵ Schmidt , T. S. B. , Raes , J. & Bork , P . The Human Gut Microbiome: From Association to Modulation . Cell 172 , 1198 – 1215 ( 2018 ). doi: 10.1016/j.cell.2018.02.044 OpenUrl CrossRef PubMed ↵ Zhang , X. , Li , L. , Butcher , J. , Stintzi , A. & Figeys , D . Advancing functional and translational microbiome research using meta-omics approaches . Microbiome 7 , 154 ( 2019 ). doi: 10.1186/s40168-019-0767-6 OpenUrl CrossRef PubMed ↵ Ojala , T. , Kankuri , E. & Kankainen , M . Understanding human health through metatranscriptomics . Trends Mol Med 29 , 376 – 389 ( 2023 ). doi: 10.1016/j.molmed.2023.02.002 OpenUrl CrossRef PubMed ↵ Bharti , R. & Grimm , D. G . Current challenges and best-practice protocols for microbiome analysis . Brief Bioinform 22 , 178 – 193 ( 2021 ). doi: 10.1093/bib/bbz155 OpenUrl CrossRef PubMed ↵ Edwin , N. R. , Fitzpatrick , A. H. , Brennan , F. , Abram , F. & O’Sullivan , O . An in-depth evaluation of metagenomic classifiers for soil microbiomes . Environ Microbiome 19 , 19 ( 2024 ). doi: 10.1186/s40793-024-00561-w OpenUrl CrossRef PubMed ↵ Lobanov , V. , Gobet , A. & Joyce , A . Ecosystem-specific microbiota and microbiome databases in the era of big data . Environ Microbiome 17 , 37 ( 2022 ). doi: 10.1186/s40793-022-00433-1 OpenUrl CrossRef ↵ Liu , Y. et al. A genome and gene catalog of glacier microbiomes . Nat Biotechnol 40 , 1341 – 1348 ( 2022 ). doi: 10.1038/s41587-022-01367-2 OpenUrl CrossRef PubMed Hiseni , P. , Rudi , K. , Wilson , R. C. , Hegge , F. T. & Snipen , L . HumGut: a comprehensive human gut prokaryotic genomes collection filtered by metagenome data . Microbiome 9 , 165 ( 2021 ). doi: 10.1186/s40168-021-01114-w OpenUrl CrossRef PubMed Feng , Y. et al. Metagenome-assembled genomes and gene catalog from the chicken gut microbiome aid in deciphering antibiotic resistomes . Commun Biol 4 , 1305 ( 2021 ). doi: 10.1038/s42003-021-02827-2 OpenUrl CrossRef Kashaf , S. S. et al. Integrating cultivation and metagenomics for a multi-kingdom view of skin microbiome diversity and functions . Nature Microbiology 7 , 169 – 179 ( 2022 ). doi: 10.1038/s41564-021-01011-w OpenUrl CrossRef ↵ Almeida , A. et al. A unified catalog of 204,938 reference genomes from the human gut microbiome . Nat Biotechnol 39 , 105 – 114 ( 2021 ). doi: 10.1038/s41587-020-0603-3 OpenUrl CrossRef PubMed ↵ Ma , B. et al. A comprehensive non-redundant gene catalog reveals extensive within-community intraspecies diversity in the human vagina . Nat Commun 11 , 940 ( 2020 ). doi: 10.1038/s41467-020-14677-3 OpenUrl CrossRef ↵ Human Microbiome Project , C. Structure, function and diversity of the healthy human microbiome . Nature 486 , 207 – 214 ( 2012 ). doi: 10.1038/nature11234 OpenUrl CrossRef PubMed Web of Science ↵ France , M. , Alizadeh , M. , Brown , S. , Ma , B. & Ravel , J . Towards a deeper understanding of the vaginal microbiota . Nat Microbiol 7 , 367 – 378 ( 2022 ). doi: 10.1038/s41564-022-01083-2 OpenUrl CrossRef ↵ Brotman , R. M. et al. Bacterial vaginosis assessed by gram stain and diminished colonization resistance to incident gonococcal, chlamydial, and trichomonal genital infection . J Infect Dis 202 , 1907 – 1915 ( 2010 ). doi: 10.1086/657320 OpenUrl CrossRef PubMed ↵ Gosmann , C. et al. in Immunity Vol . 46 29 – 37 ( 2017 ). ↵ McClelland , R. S. et al. Evaluation of the association between the concentrations of key vaginal bacteria and the increased risk of HIV acquisition in African women from five cohorts: a nested case-control study . The Lancet Infectious Diseases 18 , 554 – 564 ( 2018 ). doi: 10.1016/s1473-3099(18)30058-6 OpenUrl CrossRef ↵ Fettweis , J. M. et al. The vaginal microbiome and preterm birth . Nat Med 25 , 1012 – 1021 ( 2019 ). doi: 10.1038/s41591-019-0450-2 OpenUrl CrossRef PubMed ↵ Callahan , B. J. et al. Vol. 114 ( 2017 ). ↵ Ness , R. B. et al. A cluster analysis of bacterial vaginosis-associated microflora and pelvic inflammatory disease . Am J Epidemiol 162 , 585 – 590 ( 2005 ). doi: 10.1093/aje/kwi243 OpenUrl CrossRef PubMed Web of Science ↵ Haggerty , C. L. et al. Presence and Concentrations of Select Bacterial Vaginosis-Associated Bacteria Are Associated With Increased Risk of Pelvic Inflammatory Disease . Sex Transm Dis 47 , 344 – 346 ( 2020 ). doi: 10.1097/OLQ.0000000000001164 OpenUrl CrossRef ↵ Ratter , J. M. et al. In vitro and in vivo Effects of Lactate on Metabolism and Cytokine Production of Human Primary PBMCs and Monocytes . Front Immunol 9 , 2564 ( 2018 ). doi: 10.3389/fimmu.2018.02564 OpenUrl CrossRef PubMed ↵ Anahtar , M. N. et al. Cervicovaginal bacteria are a major modulator of host inflammatory responses in the female genital tract . Immunity 42 , 965 – 976 ( 2015 ). doi: 10.1016/j.immuni.2015.04.019 OpenUrl CrossRef PubMed ↵ Ma , B. , Forney , L. J. & Ravel, J. in Annual Review of Microbiology Vol . 66 371 – 389 ( 2012 ). ↵ Ravel , J. , et al. in Proceedings of the National Academy of Sciences Vol. 108 4680 – 4687 ( 2011 ). OpenUrl Abstract / FREE Full Text ↵ Amsel , R. et al. Nonspecific Vaginitis: Diagnostic criteria and microbial and epidemiological associations . Clinical Studies 74 , 14 – 22 ( 1983 ). OpenUrl ↵ Koumans , E. H. et al. The prevalence of bacterial vaginosis in the United States, 2001- 2004; associations with symptoms, sexual behaviors, and reproductive health . Sex Transm Dis 34 , 864 – 869 ( 2007 ). doi: 10.1097/OLQ.0b013e318074e565 OpenUrl CrossRef PubMed Web of Science ↵ Muzny , C. A. & Schwebke , J. R . Asymptomatic Bacterial Vaginosis: To Treat or Not to Treat? Curr Infect Dis Rep 22 ( 2020 ). doi: 10.1007/s11908-020-00740-z OpenUrl CrossRef Gibbs , R. S . Asymptomatic bacterial vaginosis: is it time to treat? Am J Obstet Gynecol 196 , 495 – 496 ( 2007 ). doi: 10.1016/j.ajog.2007.04.001 OpenUrl CrossRef PubMed ↵ McKinnon , L. R. et al. The Evolving Facets of Bacterial Vaginosis: Implications for HIV Transmission . AIDS Res Hum Retroviruses 35 , 219 – 228 ( 2019 ). doi: 10.1089/AID.2018.0304 OpenUrl CrossRef PubMed ↵ Gajer , P. , et al. in Science Translational Medicine Vol. 4 132ra152 ( 2012 ). OpenUrl CrossRef Munoz , A. et al. Modeling the temporal dynamics of cervicovaginal microbiota identifies targets that may promote reproductive health . Microbiome 9 , 163 ( 2021 ). doi: 10.1186/s40168-021-01096-9 OpenUrl CrossRef ↵ Srinivasan , S. et al. Temporal variability of human vaginal bacteria and relationship with bacterial vaginosis . PLoS One 5 , e10197 ( 2010 ). doi: 10.1371/journal.pone.0010197 OpenUrl CrossRef PubMed ↵ France , M. T. , et al. VALENCIA: a nearest centroid classification method for vaginal microbial communities based on composition . Microbiome 8 , 166 ( 2020 ). doi: 10.1186/s40168-020-00934-6 OpenUrl CrossRef PubMed ↵ Novak , J. et al. Characteristics associated with Lactobacillus iners-dominated vaginal microbiota . Sex Transm Infect ( 2021 ). doi: 10.1136/sextrans-2020-054824 OpenUrl Abstract / FREE Full Text van Houdt , R. et al. Lactobacillus iners-dominated vaginal microbiota is associated with increased susceptibility to Chlamydia trachomatis infection in Dutch women: a case-control study . Sex Transm Infect 94 , 117 – 123 ( 2018 ). doi: 10.1136/sextrans-2017-053133 OpenUrl Abstract / FREE Full Text Edwards , V. L. et al. The Cervicovaginal Microbiota-Host Interaction Modulates Chlamydia trachomatis Infection . mBio 10 ( 2019 ). doi: 10.1128/mBio.01548-19 OpenUrl Abstract / FREE Full Text ↵ Holm , J. B. , Carter , K. A. , Ravel , J. & Brotman , R. M . Lactobacillus iners and Genital Health: Molecular Clues to an Enigmatic Vaginal Species . Curr Infect Dis Rep 25 , 67 – 75 ( 2023 ). doi: 10.1007/s11908-023-00798-5 OpenUrl CrossRef PubMed ↵ Holm , J. B. et al. Integrating compositional and functional content to describe vaginal microbiomes in health and disease . Microbiome 11 , 259 ( 2023 ). doi: 10.1186/s40168-023-01692-x OpenUrl CrossRef PubMed ↵ Price , J. T. et al. HIV-associated vaginal microbiome and inflammation predict spontaneous preterm birth in Zambia . Sci Rep 12 , 8573 ( 2022 ). doi: 10.1038/s41598-022-12424-w OpenUrl CrossRef ↵ Deng , Z. L. et al. Metatranscriptome Analysis of the Vaginal Microbiota Reveals Potential Mechanisms for Protection against Metronidazole in Bacterial Vaginosis . mSphere 3 ( 2018 ). doi: 10.1128/mSphereDirect.00262-18 OpenUrl Abstract / FREE Full Text ↵ France , M. T. et al. Insight into the ecology of vaginal bacteria through integrative analyses of metagenomic and metatranscriptomic data . Genome Biol 23 , 66 ( 2022 ). doi: 10.1186/s13059-022-02635-9 OpenUrl CrossRef PubMed ↵ Zhu , M. et al. Vaginal Lactobacillus fatty acid response mechanisms reveal a metabolite-targeted strategy for bacterial vaginosis treatment . Cell 187 , 5413 – 5430 e5429 ( 2024 ). doi: 10.1016/j.cell.2024.07.029 OpenUrl CrossRef PubMed ↵ Bloom , S. M. et al. Cysteine dependence of Lactobacillus iners is a potential therapeutic target for vaginal microbiota modulation . Nat Microbiol 7 , 434 – 450 ( 2022 ). doi: 10.1038/s41564-022-01070-7 OpenUrl CrossRef PubMed ↵ Ravel , J. et al. 1 – 6 ( 2013 ). Elovitz , M. A. et al. Cervicovaginal microbiota and local immune response modulate the risk of spontaneous preterm delivery . Nat Commun 10 , 1305 ( 2019 ). doi: 10.1038/s41467-019-09285-9 OpenUrl CrossRef PubMed ↵ Carter , K. et al. Sexual transmission of urogenital bacteria: whole metagenome sequencing evidence from a sexual network study . mSphere 9 , e00030 – 00024 ( 2024 ). OpenUrl DiGiulio , D. B. et al. Temporal and spatial variation of the human microbiota during pregnancy . Proc Natl Acad Sci U S A 112 , 11060 – 11065 ( 2015 ). doi: 10.1073/pnas.1502875112 OpenUrl Abstract / FREE Full Text Feehily , C. et al. Shotgun sequencing of the vaginal microbiome reveals both a species and functional potential signature of preterm birth . NPJ Biofilms Microbiomes 6 , 50 ( 2020 ). doi: 10.1038/s41522-020-00162-8 OpenUrl CrossRef Yang , Q. et al. The Alterations of Vaginal Microbiome in HPV16 Infection as Identified by Shotgun Metagenomic Sequencing . Front Cell Infect Microbiol 10 , 286 ( 2020 ). doi: 10.3389/fcimb.2020.00286 OpenUrl CrossRef France , M. T. , Brown , S. E. , Rompalo , A. M. , Brotman , R. M. & Ravel , J . Identification of shared bacterial strains in the vaginal microbiota of related and unrelated reproductive-age mothers and daughters using genome-resolved metagenomics . PLoS One 17 , e0275908 ( 2022 ). doi: 10.1371/journal.pone.0275908 OpenUrl CrossRef Castillo , M. C. et al. The Zambian Preterm Birth Prevention Study (ZAPPS): Cohort characteristics at enrollment . Gates Open Res 2 , 25 ( 2018 ). doi: 10.12688/gatesopenres.12820.3 OpenUrl CrossRef Bommana , S. et al. Metagenomic shotgun sequencing of endocervical, vaginal, and rectal samples among Fijian women with and without Chlamydia trachomatis reveals disparate microbial populations and function across anatomic sites: a pilot study . Microbiology Spectrum 10 , e00105 – 00122 ( 2022 ). doi: 10.1128/spectrum.00105-22 OpenUrl CrossRef ↵ Murphy , M. S. Q. , et al. Incidental screen positive findings in a prospective cohort study in Matlab, Bangladesh: insights into expanded newborn screening for low-resource settings . Orphanet J Rare Dis 14 , 25 ( 2019 ). doi: 10.1186/s13023-018-0993-1 OpenUrl CrossRef ↵ Srinivasan , S. et al. Amygdalobacter indicium gen. nov., sp. nov., and Amygdalobacter nucleatus sp. nov., gen. nov.: novel bacteria from the family Oscillospiraceae isolated from the female genital tract . Int J Syst Evol Microbiol 73 ( 2023 ). doi: 10.1099/ijsem.0.006017 OpenUrl CrossRef ↵ Vaneechoutte , M. et al. Emended description of Gardnerella vaginalis and description of Gardnerella leopoldii sp. nov., Gardnerella piotii sp. nov. and Gardnerella swidsinskii sp. nov., with delineation of 13 genomic species within the genus Gardnerella . Int J Syst Evol Microbiol 69 , 679 – 687 ( 2019 ). doi: 10.1099/ijsem.0.003200 OpenUrl CrossRef PubMed ↵ Cantalapiedra , C. P. , Hernandez-Plaza , A. , Letunic , I. , Bork , P. & Huerta-Cepas , J . eggNOG-mapper v2: Functional Annotation, Orthology Assignments, and Domain Prediction at the Metagenomic Scale . Mol Biol Evol 38 , 5825 – 5829 ( 2021 ). doi: 10.1093/molbev/msab293 OpenUrl CrossRef PubMed ↵ Toronen , P. , Medlar , A. & Holm , L . PANNZER2: a rapid functional annotation web server . Nucleic Acids Res 46 , W84 – W88 ( 2018 ). doi: 10.1093/nar/gky350 OpenUrl CrossRef PubMed ↵ Feldgarden , M. et al. AMRFinderPlus and the Reference Gene Catalog facilitate examination of the genomic links among antimicrobial resistance, stress response, and virulence . Sci Rep 11 , 12728 ( 2021 ). doi: 10.1038/s41598-021-91456-0 OpenUrl CrossRef PubMed ↵ Huerta-Cepas , J. et al. eggNOG 4.5: a hierarchical orthology framework with improved functional annotations for eukaryotic, prokaryotic and viral sequences . Nucleic Acids Res 44 , D286 – 293 ( 2016 ). doi: 10.1093/nar/gkv1248 OpenUrl CrossRef PubMed ↵ Galperin , M. Y. , Makarova , K. S. , Wolf , Y. I. & Koonin , E. V. in Nucleic Acids Research Vol. 43 D261 – D269 ( 2015 ). OpenUrl CrossRef PubMed ↵ Finn , R. D. et al. The Pfam protein families database: towards a more sustainable future . Nucleic Acids Res 44 , D279 – 285 ( 2016 ). doi: 10.1093/nar/gkv1344 OpenUrl CrossRef PubMed ↵ Ashburner , M. et al. Gene Ontology: tool for the unification of biology . Nature Genetics 25 , 25 – 29 ( 2000 ). doi: 10.1038/75556 OpenUrl CrossRef PubMed Web of Science ↵ Biochemistry , I. U. o. & Webb , E. C. Enzyme nomenclature, 1992 : recommendations of the Nomenclature Committee of the International Union of Biochemistry and Molecular biology on the nomenclature and classification of enzymes . Sixth edition edn, ( Academic Press, Inc. , 1992 ). ↵ Schellenberger , J. , Park , J. O. , Conrad , T. M. & Palsson , B. Ø . BiGG: a Biochemical Genetic and Genomic knowledgebase of large scale metabolic reconstructions . BMC Bioinformatics 11 , 213 ( 2010 ). doi: 10.1186/1471-2105-11-213 OpenUrl CrossRef PubMed ↵ Kanehisa , M. & Goto , S . KEGG: Kyoto Encyclopedia of Genes and Genomes . Nucleic Acids Research 28 , 27 – 30 ( 2000 ). doi: 10.1093/nar/28.1.27 OpenUrl CrossRef PubMed Web of Science ↵ Drula , E. et al. The carbohydrate-active enzyme database: functions and literature . Nucleic Acids Res 50 , D571 – D577 ( 2022 ). doi: 10.1093/nar/gkab1045 OpenUrl CrossRef PubMed ↵ Blin , K. et al. antiSMASH 7.0: new and improved predictions for detection, regulation, chemical structures and visualisation . Nucleic Acids Res 51 , W46 – W50 ( 2023 ). doi: 10.1093/nar/gkad344 OpenUrl CrossRef PubMed ↵ H. Tettelin D. Medini Ma , B. , France , M. & Ravel , J. in The Pangenome: Diversity, Dynamics, and Evolution of Genomes (eds H. Tettelin & D. Medini ) ( Springer Nature Switzerland AG , 2020 ). ↵ Huang , H. , Wang , Y. , Rudin , C. & Browne , E. P . Towards a comprehensive evaluation of dimension reduction methods for transcriptomic data visualization . Commun Biol 5 , 719 ( 2022 ). doi: 10.1038/s42003-022-03628-x OpenUrl CrossRef ↵ Sousa , M. et al. Gardnerella pickettii sp. nov. (formerly Gardnerella genomic species 3) and Gardnerella greenwoodii sp. nov. (formerly Gardnerella genomic species 8) isolated from female urinary microbiome . Int J Syst Evol Microbiol 73 ( 2023 ). doi: 10.1099/ijsem.0.006140 OpenUrl CrossRef ↵ Berman , H. L. , Goltsman , D. S. , Anderson , M. , Relman , D. A. & Callahan , B. J . Gardnerella diversity and ecology in pregnancy and preterm birth . mSystems 9 , e01339 – 01323 ( 2024 ). OpenUrl ↵ Piot , P. et al. Biotypes of Gardnerella vaginalis . Journal of Clinical Microbiology 20 , 677 – 679 ( 1984 ). OpenUrl Abstract / FREE Full Text ↵ Elton , C . Competition and the structure of ecological communities . Journal of Animal Ecology 15 , 54 – 68 ( 1946 ). OpenUrl CrossRef Web of Science ↵ Brochet , S. et al. Niche partitioning facilitates coexistence of closely related honey bee gut bacteria . Elife 10 ( 2021 ). doi: 10.7554/eLife.68583 OpenUrl CrossRef PubMed ↵ Zhou , J. et al. in Applied and Environmental Microbiology , Vol. Vol. 68 326 – 334 ( 2002 ). OpenUrl Abstract / FREE Full Text ↵ Chang , C. , Bajic , D. , Vila , J. , Estrella , S. & Sanchez , A . Emergent coexistence in multispecies microbial communities . Science 381 , 343 – 348 ( 2023 ). OpenUrl CrossRef PubMed ↵ Symul , L. et al. Sub-communities of the vaginal microbiota in pregnant and non-pregnant women . Proc Biol Sci 290 , 20231461 ( 2023 ). doi: 10.1098/rspb.2023.1461 OpenUrl CrossRef PubMed ↵ Fredricks , D. N. , Fiedler , T. L. & Marrazzo , J. M. in The New England journal of medicine Vol. 353 1899 – 1911 ( 2005 ). OpenUrl CrossRef PubMed Web of Science ↵ Srinivasan , S. et al. Bacterial communities in women with bacterial vaginosis: high resolution phylogenetic analyses reveal relationships of microbiota to clinical criteria . PLoS One 7 , e37818 ( 2012 ). doi: 10.1371/journal.pone.0037818 OpenUrl CrossRef PubMed ↵ Carter , K. A. et al. Associations Between Vaginal Bacteria and Bacterial Vaginosis Signs and Symptoms: A Comparative Study of Kenyan and American Women . Front Cell Infect Microbiol 12 , 801770 ( 2022 ). doi: 10.3389/fcimb.2022.801770 OpenUrl CrossRef Srinivasan , S. et al. More Easily Cultivated Than Identified: Classical Isolation With Molecular Identification of Vaginal Bacteria . J Infect Dis 214 Suppl 1 , S21 – 28 ( 2016 ). doi: 10.1093/infdis/jiw192 OpenUrl CrossRef PubMed ↵ Srinivasan , S. et al. Vaginal Bacteria and Proinflammatory Host Immune Mediators as Biomarkers of Human Immunodeficiency Virus Acquisition Risk Among African Women . J Infect Dis ( 2024 ). doi: 10.1093/infdis/jiae406 OpenUrl CrossRef ↵ Winter , S. E. et al. Systems biology elucidates the distinctive metabolic niche filled by the human gut microbe Eggerthella lenta . PLOS Biology 21 ( 2023 ). doi: 10.1371/journal.pbio.3002125 OpenUrl CrossRef ↵ Baldewijns , S. et al. The Role of Fatty Acid Metabolites in Vaginal Health and Disease: Application to Candidiasis . Front Microbiol 12 , 705779 ( 2021 ). doi: 10.3389/fmicb.2021.705779 OpenUrl CrossRef ↵ Zini , A. , O’Bryan , M. & Schlegel , P . Nitric oxide synthase activity in human seminal plasma Urology 58 , 85 – 89 ( 2001 ). doi: 10.1016/S0090-4295(01)01001-9 OpenUrl CrossRef PubMed Web of Science ↵ Drell , T. et al. Characterization of the vaginal micro- and mycobiome in asymptomatic reproductive-age Estonian women . PLoS One 8 , e54379 ( 2013 ). doi: 10.1371/journal.pone.0054379 OpenUrl CrossRef PubMed Zhao , C. et al. Mycobiome Study Reveals Different Pathogens of Vulvovaginal Candidiasis Shape Characteristic Vaginal Bacteriome . Microbiology Spectrum 11 , e03152 – 03122 ( 2023 ). doi: 10.1128/spectrum.03152-22 OpenUrl CrossRef ↵ Bradford , L. L. & Ravel , J . The vaginal mycobiome: A contemporary perspective on fungi in women’s health and diseases . Virulence 8 , 342 – 351 ( 2017 ). doi: 10.1080/21505594.2016.1237332 OpenUrl CrossRef ↵ Armstrong , E. et al. Vaginal fungi are associated with treatment-induced shifts in the vaginal microbiota and with a distinct genital immune profile . Microbiology Spectrum 12 , e03501 – 03523 ( 2024 ). doi: 10.1128/spectrum.03501-23 OpenUrl CrossRef ↵ Liu , N.-N. et al. Mycobiome Dysbiosis in Women with Intrauterine Adhesions . Microbiology Spectrum 10 , e01324 – 01322 ( 2022 ). doi: 10.1128/spectrum.01324-22 OpenUrl CrossRef ↵ Frau , A. et al. DNA extraction and amplicon production strategies deeply inf luence the outcome of gut mycobiome studies . Sci Rep 9 , 9328 ( 2019 ). doi: 10.1038/s41598-019-44974-x OpenUrl CrossRef PubMed ↵ Akinbiyi , A. A. , Watson , R. & Feyi-Waboso , P . Prevalence of Candida albicans and bacterial vaginosis in asymptomatic pregnant women in South Yorkshire , United Kingdom. Outcome of a prospective study. Arch Gynecol Obstet 278 , 463 – 466 ( 2008 ). doi: 10.1007/s00404-008-0593-8 OpenUrl CrossRef PubMed ↵ Cotch , M. F. , Hillier , S. L. , Gibbs , R. S. & Eschenbach , D. A . Epidemiology and outcomes associated with moderate to heavy Candida colonization during pregnancy . Am J Obstet Gynecol 178 , 374 – 380 ( 1998 ). doi: 10.1016/S0002-9378(98)80028-8 OpenUrl CrossRef PubMed Web of Science ↵ Zozaya-hinchliffe , M. , Lillis , R. , Martin , D. H. & Ferris , M. J. in Journal of Clinical Microbiology Vol. 48 1812 – 1819 ( 2010 ). OpenUrl Abstract / FREE Full Text ↵ Brown , S. E. et al. in Sexually Transmitted Diseases Vol . 46 753 – 758 ( 2019 ). ↵ Eastment , M. C. et al. Association Between Vaginal Bacterial Microbiota and Vaginal Yeast Colonization . J Infect Dis 223 , 914 – 923 ( 2021 ). doi: 10.1093/infdis/jiaa459 OpenUrl CrossRef PubMed ↵ Sobel , J. D. & Vempati , Y. S . Bacterial Vaginosis and Vulvovaginal Candidiasis Pathophysiologic Interrelationship . Microorganisms 12 ( 2024 ). doi: 10.3390/microorganisms12010108 OpenUrl CrossRef ↵ Madere , F. S. et al. Transkingdom Analysis of the Female Reproductive Tract Reveals Bacteriophages form Communities . Viruses 14 ( 2022 ). doi: 10.3390/v14020430 OpenUrl CrossRef ↵ Jakobsen , R. R. et al. Characterization of the Vaginal DNA Virome in Health and Dysbiosis . Viruses 12 ( 2020 ). doi: 10.3390/v12101143 OpenUrl CrossRef ↵ Blackwell , A. L. Vaginal bacterial phaginosis . Sex Transm Infect 75 , 352 – 353 ( 1999 ). OpenUrl CrossRef PubMed Web of Science ↵ Henrot , C. & Petit , M. A . Signals triggering prophage induction in the gut microbiota . Mol Microbiol 118 , 494 – 502 ( 2022 ). doi: 10.1111/mmi.14983 OpenUrl CrossRef PubMed ↵ Happel , A. U. et al. In Silico Characterisation of Putative Prophages in Lactobacillaceae Used in Probiotics for Vaginal Health . Microorganisms 10 ( 2022 ). doi: 10.3390/microorganisms10020214 OpenUrl CrossRef Happel , A. U. et al. Exploring potential of vaginal Lactobacillus isolates from South African women for enhancing treatment for bacterial vaginosis . PLoS Pathog 16 , e1008559 ( 2020 ). doi: 10.1371/journal.ppat.1008559 OpenUrl CrossRef ↵ Martin , R. , Escobedo , S. & Suarez , J. E . Induction, structural characterization, and genome sequence of Lv1, a prophage from a human vaginal Lactobacillus jensenii strain . Int Microbiol 13 , 113 – 121 ( 2010 ). doi: 10.2436/20.1501.01.116 OpenUrl CrossRef PubMed ↵ Colombini , L. et al. The mobilome of Lactobacillus crispatus M247 includes two novel genetic elements: Tn7088 coding for a putative bacteriocin and the siphovirus prophage PhiM247 . Microb Genom 9 ( 2023 ). doi: 10.1099/mgen.0.001150 OpenUrl CrossRef ↵ Cornuault , J. K. et al. The enemy from within: a prophage of Roseburia intestinalis systematically turns lytic in the mouse gut, driving bacterial adaptation by CRISPR spacer acquisition . ISME J 14 , 771 – 787 ( 2020 ). doi: 10.1038/s41396-019-0566-x OpenUrl CrossRef PubMed ↵ Lossouarn , J. et al. Enterococcus faecalis Countermeasures Defeat a Virulent Picovirinae Bacteriophage . Viruses 11 ( 2019 ). doi: 10.3390/v11010048 OpenUrl CrossRef ↵ Zeng , Z. et al. Cold adaptation regulated by cryptic prophage excision in Shewanella oneidensis . ISME J 10 , 2787 – 2800 ( 2016 ). doi: 10.1038/ismej.2016.85 OpenUrl CrossRef PubMed ↵ Wang , W. et al. The coral pathogen Vibrio coralliilyticus kills non-pathogenic holobiont competitors by triggering prophage induction . Nat Ecol Evol 6 , 1132 – 1144 ( 2022 ). doi: 10.1038/s41559-022-01795-y OpenUrl CrossRef PubMed ↵ James , C. E. et al. in The ISME Journal Vol. 9 1391 – 1398 ( Nature Publishing Group , 2014 ). OpenUrl ↵ Borodovich , T. , Shkoporov , A. N. , Ross , R. P. & Hill , C . Phage-mediated horizontal gene transfer and its implications for the human gut microbiome . Gastroenterol Rep (Oxf ) 10 , goac012 ( 2022 ). doi: 10.1093/gastro/goac012 OpenUrl CrossRef ↵ France , M. et al. Insight into the ecology of vaginal bacteria through integrative analyses of metagenomic and metatranscriptomic data . Zenodo ( 2022 ). doi: 10.5281/zenodo.5932449 OpenUrl CrossRef ↵ Pelayo , P. et al. Prevotella are major contributors of sialidases in the human vaginal microbiome . Proc Natl Acad Sci U S A 121 , e2400341121 ( 2024 ). doi: 10.1073/pnas.2400341121 OpenUrl CrossRef PubMed ↵ Jenkins , D. J. et al. Bacterial amylases enable glycogen degradation by the vaginal microbiome . Nat Microbiol 8 , 1641 – 1652 ( 2023 ). doi: 10.1038/s41564-023-01447-2 OpenUrl CrossRef ↵ Zhang , Q. Q. et al. Prebiotic Maltose Gel Can Promote the Vaginal Microbiota From BV- Related Bacteria Dominant to Lactobacillus in Rhesus Macaque . Front Microbiol 11 , 594065 ( 2020 ). doi: 10.3389/fmicb.2020.594065 OpenUrl CrossRef ↵ Chen , S. , Zhou , Y. , Chen , Y. & Gu , J. fastp: an ultra-fast all-in-one FASTQ preprocessor . Bioinformatics 34 , i884 – i890 ( 2018 ). doi: 10.1093/bioinformatics/bty560 OpenUrl CrossRef PubMed ↵ Nurk , S. , Meleshko , D. , Korobeynikov , A. & Pevzner , P. A . metaSPAdes: a new versatile metagenomic assembler . Genome Res 27 , 824 – 834 ( 2017 ). doi: 10.1101/gr.213959.116 OpenUrl Abstract / FREE Full Text ↵ Wu , Y. W. , Simmons , B. A. & Singer , S. W . MaxBin 2.0: an automated binning algorithm to recover genomes from multiple metagenomic datasets . Bioinformatics 32 , 605 – 607 ( 2016 ). doi: 10.1093/bioinformatics/btv638 OpenUrl CrossRef PubMed ↵ Kang , D. D. et al. MetaBAT 2: an adaptive binning algorithm for robust and efficient genome reconstruction from metagenome assemblies . PeerJ 7 , e7359 ( 2019 ). doi: 10.7717/peerj.7359 OpenUrl CrossRef PubMed ↵ Liu , C. C. et al. MetaDecoder: a novel method for clustering metagenomic contigs . Microbiome 10 , 46 ( 2022 ). doi: 10.1186/s40168-022-01237-8 OpenUrl CrossRef PubMed ↵ Uritskiy , G. V. , DiRuggiero , J. & Taylor , J . MetaWRAP-a flexible pipeline for genome-resolved metagenomic data analysis . Microbiome 6 , 158 ( 2018 ). doi: 10.1186/s40168-018-0541-1 OpenUrl CrossRef PubMed ↵ Parks , D. H. , Imelfort , M. , Skennerton , C. T. , Hugenholtz , P. & Tyson , G. W . CheckM: assessing the quality of microbial genomes recovered from isolates, single cells, and metagenomes . Genome Res 25 , 1043 – 1055 ( 2015 ). doi: 10.1101/gr.186072.114 OpenUrl Abstract / FREE Full Text ↵ Nayfach , S. , Shi , Z. J. , Seshadri , R. , Pollard , K. S. & Kyrpides , N. C . New insights from uncultivated genomes of the global human gut microbiome . Nature 568 , 505 – 510 ( 2019 ). doi: 10.1038/s41586-019-1058-x OpenUrl CrossRef PubMed ↵ Hyatt , D. et al. Prodigal: prokaryotic gene recognition and translation initiation site identfication . BMC Bioinformatics 11 ( 2010 ). doi: 10.1186/1471-2105-11-119 OpenUrl CrossRef PubMed ↵ Li , W. & Godzik , A . Cd-hit: a fast program for clustering and comparing large sets of protein or nucleotide sequences . Bioinformatics 22 , 1658 – 1659 ( 2006 ). doi: 10.1093/bioinformatics/btl158 OpenUrl CrossRef PubMed Web of Science ↵ Nurk , S. et al. The complete sequence of a human genome . Science 376 , 44 – 53 ( 2022 ). doi: 10.1126/science.abj6987 OpenUrl CrossRef PubMed ↵ Wu , F. et al. A new coronavirus associated with human respiratory disease in China . Nature 579 , 265 – 269 ( 2020 ). doi: 10.1038/s41586-020-2008-3 OpenUrl CrossRef PubMed ↵ Camacho , C. et al. BLAST+: architecture and applications . BMC Bioinformatics 10 , 421 ( 2009 ). doi: 10.1186/1471-2105-10-421 OpenUrl CrossRef PubMed ↵ Chaumeil , P. A. , Mussig , A. J. , Hugenholtz , P. & Parks , D. H . GTDB-Tk: a toolkit to classify genomes with the Genome Taxonomy Database . Bioinformatics 36 , 1925 – 1927 ( 2019 ). doi: 10.1093/bioinformatics/btz848 OpenUrl CrossRef PubMed ↵ Tatusov , R. L. , Koonin , E. V. & Lipman , D. J . A Genomic Perspective on Protein Families . Science 278 , 631 – 637 ( 1997 ). doi: 10.1126/science.278.5338.631 OpenUrl Abstract / FREE Full Text ↵ Gene Ontology , C. , et al. The Gene Ontology knowledgebase in 2023 . Genetics 224 ( 2023 ). doi: 10.1093/genetics/iyad031 OpenUrl CrossRef PubMed ↵ Bairoch , A . The ENZYME database in 2000 . Nucleic Acids Research 28 , 304 – 305 ( 2000 ). doi: 10.1093/nar/28.1.304 OpenUrl CrossRef PubMed Web of Science ↵ Emms , D. M. & Kelly , S . OrthoFinder: phylogenetic orthology inference for comparative genomics . Genome Biol 20 , 238 ( 2019 ). doi: 10.1186/s13059-019-1832-y OpenUrl CrossRef PubMed ↵ Guo , J. et al. VirSorter2: a multi-classifier, expert-guided approach to detect diverse DNA and RNA viruses . Microbiome 9 , 37 ( 2021 ). doi: 10.1186/s40168-020-00990-y OpenUrl CrossRef PubMed ↵ Nayfach , S. et al. CheckV assesses the quality and completeness of metagenome-assembled viral genomes . Nat Biotechnol 39 , 578 – 585 ( 2021 ). doi: 10.1038/s41587-020-00774-7 OpenUrl CrossRef PubMed ↵ Shang , J. , Tang , X. & Sun , Y . PhaTYP: predicting the lifestyle for bacteriophages using BERT . Brief Bioinform 24 ( 2023 ). doi: 10.1093/bib/bbac487 OpenUrl CrossRef ↵ Shang , J. , Jiang , J. & Sun , Y . Bacteriophage classification for assembled contigs using graph convolutional network . Bioinformatics 37 , i25 – i33 ( 2021 ). doi: 10.1093/bioinformatics/btab293 OpenUrl CrossRef PubMed ↵ Lu , C. et al. Prokaryotic virus host predictor: a Gaussian model for host prediction of prokaryotic viruses in metagenomics . BMC Biol 19 , 5 ( 2021 ). doi: 10.1186/s12915-020-00938-6 OpenUrl CrossRef ↵ Sieber , C. M. K. et al. Recovery of genomes from metagenomes via a dereplication, aggregation and scoring strategy . Nat Microbiol 3 , 836 – 843 ( 2018 ). doi: 10.1038/s41564-018-0171-1 OpenUrl CrossRef ↵ Gourle , H. , Karlsson-Lindsjo , O. , Hayer , J. & Bongcam-Rudloff , E . Simulating Illumina metagenomic data with InSilicoSeq . Bioinformatics 35 , 521 – 522 ( 2019 ). doi: 10.1093/bioinformatics/bty630 OpenUrl CrossRef PubMed ↵ Lin , H. & Peddada , S. D . Analysis of compositions of microbiomes with bias correction . Nat Commun 11 , 3514 ( 2020 ). doi: 10.1038/s41467-020-17041-7 OpenUrl CrossRef PubMed ↵ Lin , H. , Eggesbo , M. & Peddada , S. D . Linear and nonlinear correlation estimators unveil undescribed taxa interactions in microbiome data . Nat Commun 13 , 4946 ( 2022 ). doi: 10.1038/s41467-022-32243-x OpenUrl CrossRef PubMed ↵ van Heel , A. J. et al. BAGEL4: a user-friendly web server to thoroughly mine RiPPs and bacteriocins . Nucleic Acids Res 46 , W278 – W281 ( 2018 ). doi: 10.1093/nar/gky383 OpenUrl CrossRef PubMed ↵ Bouras , G. et al. Pharokka: a fast scalable bacteriophage annotation tool . Bioinformatics 39 ( 2023 ). doi: 10.1093/bioinformatics/btac776 OpenUrl CrossRef PubMed ↵ Hoffman , G. E. & Schadt , E. E . variancePartition: interpreting drivers of variation in complex gene expression studies . BMC Bioinformatics 17 , 483 ( 2016 ). doi: 10.1186/s12859-016-1323-z OpenUrl CrossRef PubMed ↵ Kanehisa , M. , Sato , Y. & Morishima , K . BlastKOALA and GhostKOALA: KEGG Tools for Functional Characterization of Genome and Metagenome Sequences . J Mol Biol 428 , 726 – 731 ( 2016 ). doi: 10.1016/j.jmb.2015.11.006 OpenUrl CrossRef PubMed View the discussion thread. Back to top Previous Next Posted March 04, 2025. Download PDF Supplementary Material Data/Code Email Thank you for your interest in spreading the word about bioRxiv. NOTE: Your email address is requested solely to identify you as the sender of this article. Your Email * Your Name * Send To * Enter multiple addresses on separate lines or separate them with commas. You are going to email the following VIRGO2: Unveiling the Functional and Ecological Complexity of the Vaginal Microbiome with an Enhanced Non-Redundant Gene Catalog Message Subject (Your Name) has forwarded a page to you from bioRxiv Message Body (Your Name) thought you would like to see this page from the bioRxiv website. Your Personal Message CAPTCHA This question is for testing whether or not you are a human visitor and to prevent automated spam submissions. Share VIRGO2: Unveiling the Functional and Ecological Complexity of the Vaginal Microbiome with an Enhanced Non-Redundant Gene Catalog M. T. France , I. Chaudry , L. Rutt , M. Quain , B. Shirtliff , E. McComb , A. Maros , M. Alizadeh , F. A. Hussain , M. A. Elovitz , D. A. Relman , A. Rahman , R. M. Brotman , J. Price , M. Kassaro , J. B. Holm , B. Ma , J. Ravel bioRxiv 2025.03.04.641479; doi: https://doi.org/10.1101/2025.03.04.641479 Share This Article: Copy Citation Tools VIRGO2: Unveiling the Functional and Ecological Complexity of the Vaginal Microbiome with an Enhanced Non-Redundant Gene Catalog M. T. France , I. Chaudry , L. Rutt , M. Quain , B. Shirtliff , E. McComb , A. Maros , M. Alizadeh , F. A. Hussain , M. A. Elovitz , D. A. Relman , A. Rahman , R. M. Brotman , J. Price , M. Kassaro , J. B. Holm , B. Ma , J. Ravel bioRxiv 2025.03.04.641479; doi: https://doi.org/10.1101/2025.03.04.641479 Citation Manager Formats BibTeX Bookends EasyBib EndNote (tagged) EndNote 8 (xml) Medlars Mendeley Papers RefWorks Tagged Ref Manager RIS Zotero Tweet Widget Facebook Like Google Plus One Subject Area Microbiology Subject Areas All Articles Animal Behavior and Cognition (7635) Biochemistry (17697) Bioengineering (13895) Bioinformatics (41951) Biophysics (21456) Cancer Biology (18594) Cell Biology (25520) Clinical Trials (138) Developmental Biology (13381) Ecology (19903) Epidemiology (2067) Evolutionary Biology (24323) Genetics (15612) Genomics (22510) Immunology (17737) Microbiology (40401) Molecular Biology (17183) Neuroscience (88622) Paleontology (667) Pathology (2833) Pharmacology and Toxicology (4825) Physiology (7644) Plant Biology (15158) Scientific Communication and Education (2046) Synthetic Biology (4296) Systems Biology (9825) Zoology (2271)