- Open Access
TypOn: the microbial typing ontology
© Vaz et al.; licensee BioMed Central Ltd. 2014
Received: 20 June 2014
Accepted: 6 October 2014
Published: 18 October 2014
Bacterial identification and characterization at subspecies level is commonly known as Microbial Typing. Currently, these methodologies are fundamental tools in Clinical Microbiology and bacterial population genetics studies to track outbreaks and to study the dissemination and evolution of virulence or pathogenicity factors and antimicrobial resistance. Due to advances in DNA sequencing technology, these methods have evolved to become focused on sequence-based methodologies. The need to have a common understanding of the concepts described and the ability to share results within the community at a global level are increasingly important requisites for the continued development of portable and accurate sequence-based typing methods, especially with the recent introduction of Next Generation Sequencing (NGS) technologies. In this paper, we present an ontology designed for the sequence-based microbial typing field, capable of describing any of the sequence-based typing methodologies currently in use and being developed, including novel NGS based methods. This is a fundamental step to accurately describe, analyze, curate, and manage information for microbial typing based on sequence based typing methods.
It is widely known that different strains from a given bacterial species may have distinct phenotypic behaviors, such as a higher capacity to cause invasive disease, to asymptomatically colonize the host or to present resistance to antimicrobials . Such distinguishing characteristics can be usually attributed to lineages identified at the level of the genotype. Microbial typing refers to the methodologies used to identify these lineages and define them at sub-species level. Microbial typing has important implications in several health related fields such as surveillance of infectious diseases, outbreak investigation and control, identification of pathogen reservoirs, and studies on pathogenesis . Traditionally these methodologies were based on characterizing a limited number of markers. These markers can be phenotypic characteristics, such as the presence of certain structures on the bacterial surface , or genetic characteristics, such as the presence on the bacterial genome of DNA sequences that are recognized and cleaved by specific enzymes, generating band patterns by gel electrophoresis . More recently, due to the low cost and increasing availability of DNA sequencing technologies, the development of typing methods became focused on the use of DNA sequence information.
Although these methods revolutionized microbial typing, through the creation of novel, unambiguous and easily understandable nomenclatures for human use, the existing databases still lack interfaces for machine-readable formats, which can be used for automated data submission and querying. An ontology describing the concepts and relationships for sequence-based typing methods can thus provide a powerful tool in the field. Sharing data annotated in a common language, and in a semantically rich machine-readable format, will allow a better integration of the data existing in databases of sequence-based typing methods, epidemiological information and the novel NGS data being produced . Furthermore, it can facilitate comparison of different typing schemas, and allow users to mine, in an effective way, the ever-growing public data.
In this paper, we describe the design of TypOn, a microbial typing ontology. TypOn was developed from a previous prototype ontology , and focuses on sequence-based typing methods, including novel NGS methodologies. We discuss the connection of TypOn to existing ontologies, how to use it to annotate data already publicly available, and the methods to effectively query it.
Multilocus Sequence Typing (MLST) [8, 9] is a widely adopted methodology to type several diferent species of microorganisms. This method is based on determining the sequence of internal fragments of multiple (typically seven) loci encoding proteins with housekeeping functions. By a locus we understand a specific location in the chromosome, where different sequences occupying a given locus define distinct alleles of that locus. Each set of loci defining an MLST scheme is proposed by a group of researchers that usually also provide evidence supporting its discriminatory power and suitability for its intended purpose. When applying the methodology, the set of alleles identified at the loci considered define a sequence type (ST), a key identifier with this methodology. The loci chosen are usually different for each species, although some species may share some or even all loci in their MLST schemas. The number of loci can be also variable and can be greater or smaller than the seven loci more commonly adopted. MLST’s large appeal for the community was the reproducibility and portability of results, which allowed the deployment of databases for several bacterial species [10–12]. The strain nomenclature developed by MLST facilitated the global tracking and immediate comparison of microbial strains in clinical and research settings.
Another sequence-based method that derived its success from a common nomenclature and the ease of strain classification, was spa typing for Staphylococcus aureus, an important pathogen that is one of the major causes of nosocomial infections . This method is based on repeat sequences present at a single locus, the spa gene. These repeats are short sequences of DNA (in the case of spa about thirty nucleotides) that, although sharing consensus characteristics, can be variable in their sequence. Their number is also variable and, in the case of spa typing, this is expressed by a string of numbers representing the identity and number of repeats present at each spa allele. An updated list of identified repeats and spa types can be found at http://spaserver.ridom.de/.
Multilocus Variable Number of Tandem Repeats Analysis (MLVA)  is a method that is based on the number of repeat patterns present on several defined locus that are, similarly to MLST, defined in a schema. Several schemas are also available in multiple websites such as http://www.mlva.eu/, http://www.mlva.net/ or http://www.miru-vntrplus.org/.
With the advent of Next Generation Sequencing (NGS) technologies, and the ability to produce a draft genome sequence of a microbial strain in a couple of days instead of weeks or months, researchers can use this information to classify the strain according to new or previously available sequence-based typing methods. Furthermore, novel typing methods are being developed that are able to probe tens, hundreds or even thousands of different loci across the genome , effectively extending the MLST concepts. An example of such a method is ribosomal MLST . Other whole genome methodologies probe the genome for Single Nucleotide Polymorphisms (SNPs) when compared to reference genomes [18, 19] in order to discriminate strains.
The main goal of typing methods is the characterization of individuals existing in a given sample. The sample under study can be recovered from sick or healthy subjects or directly from the environment. The first process is the isolation of the microorganisms to be characterized from the sample collected. Each individual, or in the case of bacteria, each colony isolated from the microbial population then becomes an “isolate”, referring to the process of isolating it from a sample that contains many microorganisms potentially representing distinct species. In particular each isolate can then be identified at subspecies, species, or genus level, identifying it as belonging to a given taxonomic unit, i.e., taxon. Moreover, since most microorganisms reproduce asexually the subsequent propagation of this isolate in the laboratory as an axenic culture, would be expected to generate a clonal population.
Each isolate can be associated with typing information and ancillary details. The isolate can have the nucleotide sequence of its genetic material determined and multiple typing characteristics defined through different typing methods, such as MLST, spa typing or MLVA . There are several categories of typing methods. For instance, although MLST and spa typing are both genotypic methods, MLST is a multi-locus typing method while spa typing is a single-locus typing method.
For the specific case of multi-locus typing methods, several schemas have been defined as indicated above. Each schema is characterized by the (possibly ordered) set of loci selected for a given taxon, usually defined at the species level. Each schema is then administrated by one or a group of microbiologist, or by an institution, and each isolate’s ancillary data and respective typing information are deposited in public repositories and validated by a curator. Different schemas can be defined for the same taxon and an ontology will facilitate understanding the relationships between different schemas. Ancillary data include information about the place where the microorganism was isolated, information about environment or host, and other possible contextual details. Later on, we will discuss how this information can be added and annotated in the context of TypOn.
We also note that data ownership is a particular delicate issue in the surveillance of communicable diseases and, as we will discuss later, the approach proposed in this paper allows a straightforward implementation of the agreed upon policies. The sensitive nature of the information and the ethical issues associated can be also safeguarded through the application of suitable access control levels based on the ownership structure embedded in the ontology.
Microbial typing ontology
TypOn is an OWL ontology for describing microbial typing, focusing on sequence based methods. Such a description will provide a universal framework for the exchange of information on the microbial typing field, allowing the integration of data coming from the numerous and disparate online databases.
In the next Section, we describe and discuss the TypOn ontology and its suitability for describing knowledge in the microbiology typing methods domain. Later, we illustrate how to annotate and query microbial typing data and information using TypOn.
Ongoing developments, new versions, as well as use cases and examples, can be found in the project repository https://bitbucket.org/phyloviz/typon, and documented in the project wiki https://bitbucket.org/phyloviz/typon/wiki. The base URL for TypOn, http://purl.phyloviz.net/ontology/typon, redirects transparently to the last stable version of the ontology, in RDF/XML format, located in the master branch in the development repository.
Statistics concerning concepts and properties either defined in TypOn or reused from other ontologies/vocabularies
Ontology (BFO) 
Ontology for Biomedical
Investigations (OBI) 
Ontology (UNIPROT) 
Friend of a
Friend (FOAF) 
Although we will discuss examples of bacterial typing, we believe that TypOn is equally applicable to typing methods used to characterize any microorganism. Naturally, this implies that other loci, in addition to those used in the examples, will have to be entered into the ontology and new schemas will have to be defined. The ontology offers a flexible framework with which the existing and future sequence typing methods can be described and the examples are meant to illustrate its application as well as its flexibility. As described above, the first process in microbial typing is the recovery of the microorganisms to be characterized from the sample collected and, thus, Isolate is a main concept in TypOn and it is characterized by several properties.
Figure 1 shows an overview of the Isolate class and its related concepts and properties.
Each Isolate when identified at subspecies, species, or genus level, belongs to a certain Taxon, a relation that we express through the property isFromTaxon. The Taxon concept is reused from the Uniprot Core Ontology  for classifying life forms. Moreover, we define that each Isolate is an Organism, a concept that is reused from the Ontology for Biomedical Investigations (OBI) . We also know the Place from where each Isolate was recovered, which we describe through the property isolatedAt. As with the Taxon concept, the Place concept is reused from another ontology, in this case the DBpedia ontology . One can also define the environment material or system where a given Isolate was collected. These concepts are already found in the environment ontology (EnvO)  and, hence, we reused them and we add the properties isolatedOnMaterial and isolatedOnSystem, relating these concepts with the concept Isolate. Each Isolate, can have Sequence information, i.e. the nucleotide sequence of its genetic material, and Typing Information. The property hasTypingInformation commonly has cardinality higher than one for each Isolate, since several typing methods, such as MLST or MLVA, can be applied to the same Isolate. Later we will present an example of an isolate with both MLST and spa typing information.
In this context, it is important to note that TypingInformation is the root of a class hierarchy (see Figure 2). This hierarchy can be extended by including new, and already known, typing techniques, such as phenotypic information related to antibiotic susceptibility. In particular, we are able to distinguish different categories of typing methods. Let us consider MLST and spa Typing. Both are Genotypic methods, but the first is a Multilocus method while the latter is a Single Locus method.
As depicted in Figure 1 and indicated above, each Isolate may have been characterized by more than one typing method. In the case of MLST, this kind of typing information can be subject to different schemas, resulting in different sequence types, which are characterized by the alleles found at each locus. Therefore, in our ontology, we associate to each MLST both a schema and a set of observed alleles through properties hasSchema and hasIdentifiedAllele, respectively. The property hasSchema has cardinality of exactly one and, hence, MLST and MLVA instances must be related with one and only one given Schema. Notice that STAllele is an Allele (see Figure 1) defined in a MLST Schema, disjoint from Spa Allele and VNTAllele. So, we only associate to MLST typing information the concept of STAllele. This is necessary given the different ways in which the distinct alleles are defined in these typing methods.
The entity Spa Allele is used in the context of spa typing. Each spa typing information has a matching spa Allele, which corresponds to a specific sequence of repeats found as a result of the amplification of the locus of the spa gene of Staphylococcus aureus. In Figure 2 we can observe that a given Spa Allele has repeat profile parts, i.e., an entity that stores the index order of a given Spa Repeat in a given Spa Allele. Notice that each Spa Repeat may occur more than once in a given Spa Allele and that different Spa Alleles can have distinct number of repeats.
Notice also the difference between properties hasIdentifiedAllele and hasDefinedAllele relating respectively MLST and Locus concepts with the Allele concept. One could imagine that alleles associated to a given MLST instance could be obtained through the defined Schema, since property hasDefinedAllele allows to relate loci and alleles. But this is not the case. A Locus may have associated many alleles, with each of them belonging to many MLST instances, and hence we cannot identify the allele belonging to a particular MLST instance. That can be accomplished through property hasIdentifiedAllele which relates each MLST instance with its identified alleles. Although we did not add those kind of assertions in our ontology, we can still use this information to assert that the identified alleles for a given MLST instance are sufficient against a given Schema. We could even infer for which schemas a given MLST instance provides enough information.
It is also important to note that, by knowing only the Locus, it is possible to identify the Taxon that it belongs to, using the isOfTaxon property (see Figure 1).
Additional information for each class, such as sample collection date and other id, are described through data properties. For instance, the class Allele has data properties such as id and date entered. The class Isolate has data properties such as sample collection date and date entered.
Whenever possible, we reuse concepts from and establish relations with other ontologies as indicated in Table 1. An Agent is a concept imported from the FOAF ontology  and it can be a person, a group of persons or even an institution. In TypOn it allows the description of a person or a group of persons who have curated the information about the isolate, who have submitted that information to the database and who own the isolate. These relations are described by the object properties curated by, sent by and owned by, respectively. Reusing the Agent concept is extremely useful because it will allow, for instance, the use of TypOn together with the Web Access Control ontology  for defining access control levels in microbial typing databases, an important issue as mentioned above. Several applications in defining access control have been widely discussed and are well known to the research community [31, 32].
All TypOn concepts were derived from the Basic Formal Ontology (BFO) , the Ontology for Biomedical Investigations (OBI)  and the Sequence Ontology (SO)  to ensure upper-level interoperability with other ontologies. In order to avoid huge imports in TypOn we have used OntoFox  to query and import only relevant concepts in top level ontologies. These are then imported as, and are available at, https://bitbucket.org/phyloviz/typon/raw/master/imported.owl.
TypOn was also submitted to the BioPortal (https://bioportal.bioontology.org/), hosted by the National Center of Biomedical Ontologies (NCBO), being also available at http://bioportal.bioontology.org/ontologies/TYPON.
Another example is the Spa Repeat concept which is a repeat unit (SO:0000657). Both Locus and Spa Repeat are also generically dependent continuant (BFO _0000031) since region (SO:0000001) and repeat unit (SO:0000657) concepts are subclasses of BFO _0000031 as defined in the sequence ontology (SO). Thus these concepts are related to both SO and BFO (see Figures 4 and 5).
In this section, we illustrate how we can represent typing information annotated with the TypOn ontology. Our example makes use of data regarding the characterization of a Staphylococcus aureus isolate for the purpose of this case. We will use the Turtle language  in the description of our isolate.
In this section we discuss how to annotate a large dataset with the TypOn ontology and how to perform queries. We started by writing a D2RQ mapping for the data available for Neisseria spp, one of the databases available in our local BIGSdb  installation. D2RQ [35, 36] is a mapping language and platform for treating non-RDF relational databases as virtual RDF graphs, aiming to expose RDBs on the semantic web. The mapping allows us to reuse existing vocabularies and ontologies, such as TypOn, to map relational schemas, such as the one underlying BIGSdb. We note that our mapping is not exhaustive and that we just annotated part of the data with TypOn ontology. The mapping is available at https://bitbucket.org/phyloviz/typon/raw/master/test/BIGSdb_d2r_mapping.ttl.
Even though the D2RQ web app provides a SPARQL endpoint, it turns out that queries may take a long time to complete, causing high loads in the underlying database, and the web app may also become irresponsive. To overcome this issue we used the tool dump-rdf available with D2RQ to dump all triples and we uploaded them to a local instance of Virtuoso, which among other functionalities includes a highly efficient triple store (http://virtuoso.openlinksw.com/). Hence, a more responsive SPARQL endpoint is available at http://data.phyloviz.net/sparql, where we should select http://rest.phyloviz.net/neisseria/ as the default graph. All resources at http://rest.phyloviz.net/neisseria/ are also dereferenceable through rewrite rules against the SPARQL endpoint.
where the prefix db: stands for http://rest.phyloviz.net/neisseria/resource/.
TypOn provides the basic concepts needed to establish the vocabulary and the semantic relationships for different sequence-based typing methods, and it is designed to allow further expansion. It was defined based on three different approaches to sequence-based typing: using the DNA sequence information directly, using the sequence of repeats in a DNA sequence, and for MLVA, using the number of repeats in a locus. Since these three approaches can be used to define many of the existing typing methods, TypOn can be easily expanded to encompass the newer multilocus typing techniques that are appearing based on NGS technologies, defined by expansion of the MLST concepts to larger numbers of genes  or by Single Nucleotide Polymorphism approaches, where each position on the genome can be viewed as a locus and the nucleotide present as an allele. Other advantages of this ontology is that it can provide a consistent link with legacy microbial typing techniques and provide a way to describe and annotate the evolution of specific typing schemas. This will be of paramount importance, if schemas that will be constructed by grouping loci from existing schemas or adding new loci, are to be designed and represented in an accurate way. This ontology is the first stepping stone on the implementation of a semantic web approach for the data repositories in this field. It lays the foundation for a common language that can be used to integrate and link data from different typing databases and for a complete merging of microbial typing with microbial genomics. Using the strategy discussed in the previous section (Annotating data), a SPARQL endpoint is already deployed for the Pubmlst MLST databases at http://pubmlst.org/sparql. This endpoint accesses data annotated using TypOn for MLST databases for 75 distinct bacterial species that are hosted at Pubmlst.org and further 29 species hosted externally to Pubmlst.org. A RESTful API is also being developed to facilitate data access without requiring the SPARQL endpoint. Future work will focus on expanding the ontology and creating and deploying RESTful APIs to perform not only custom querying but also automated submission and curation of data for authenticated users, in order to speed up and distribute the curating process, and ensure better quality and reproducibility of data in the field of microbial typing.
The work presented in this paper made use of data available at MLST.net , PubMLST  and Institut Pasteur MLST Databases . This study was partly supported by the European Community grant FP7-278864-2 (PathoNgenTrace, http://www.patho-ngen-trace.eu/), and by national funds through FCT – Fundação para a Ciência e Tecnologia, under projects PTDC/EIA-CCO/118533/2010, EXCL/EEI-ESS/0257/2012, and PEst-OE/EEI/LA0021/2013.
- Sá-Leão R, Pinto F, Aguiar S, Nunes S, Carriço JA, Frazão N, Gonçalves-Sousa N, Melo-Cristino J, de Lencastre H, Ramirez M:Analysis of invasiveness of pneumococcal serotypes and clones circulating in portugal before widespread use of conjugate vaccines reveals heterogeneous behavior of clones expressing the same serotype. J Clin Microbiol. 2011, 49 (1): 1369-1375.View ArticleGoogle Scholar
- Struelens MJ:Consensus guidelines for appropriate use and evaluation of microbial epidemiologic typing systems. Clin Microbiol Infect. 1996, 2 (1): 2-11. 10.1111/j.1469-0691.1996.tb00193.x.View ArticleGoogle Scholar
- Brueggemann AB, Griffiths DT, Meats E, Peto T, Crook DW, Spratt BG:Clonal relationships between invasive and carriageStreptococcus pneumoniae, and serotype- and clone-specific differences in invasive disease potential. J Infect Dis. 2003, 187 (9): 1424-1432. 10.1086/374624.View ArticleGoogle Scholar
- Allardet-Servent A, Bouziges N, Carles-Nurit MJ, Bourg G, Gouby A, Ramuz M:Use of low-frequency-cleavage restriction endonucleases for DNA analysis in epidemiological investigations of nosocomial bacterial infections. J Clin Microbiol. 1989, 27 (9): 2057-2061.Google Scholar
- Carriço JA, Sabat AJ, Friedrich AW, Ramirez M, on behalf of the ESCMIDStudyGroupforEpidemiologicalMarkers(ESGEM):Bioinformatics in bacterial molecular epidemiology and public health: databases, tools and the next-generation sequencing revolution. Euro Surveill. 2013, 18 (4): 20382-Google Scholar
- Almeida J, Tiple J, Ramirez M, Melo-Cristino J, Vaz C, Francisco AP, Carriço JA:An ontology and a rest api for sequence based microbial typing data. In JBI’2010. Lecture Notes in Computer Science, vol. 6620. Berlin Heidelberg: Springer; 2012:21–28,Google Scholar
- Sabat AJ, Budimir A, Nashev D, Sa-Leao R, van Dijl Jm, Laurent F, Grundmann H, Friedrich AW:Overview of molecular typing methods for outbreak detection and epidemiological surveillance. Euro Surveill. 2013, 18 (4): 20380-Google Scholar
- Spratt BG:Multilocus sequence typing: molecular typing of bacterial pathogens in an era of rapid DNA sequencing and the internet. Curr Opin Microbiol. 1999, 2 (3): 312-316. 10.1016/S1369-5274(99)80054-X.MathSciNetView ArticleGoogle Scholar
- Maiden MC, Bygraves JA, Feil EJ, Morelli G, Russell JE, Urwin R, Zhang Q, Zhou J, Zurth K, Caugant DA, Feavers IM, Achtman M, Spratt BG:Multilocus sequence typing: a portable approach to the identification of clones within populations of pathogenic microorganisms. Proc Natl Acad Sci U S A. 1998, 95 (6): 3140-3145. 10.1073/pnas.95.6.3140.View ArticleGoogle Scholar
- MLST: Multi Locus Sequence Typing. [http://www.mlst.net], Imperial College of London,
- PubMLST. [http://pubmlst.org/], University of Oxford (UK),
- Institut Pasteur MLST Databases. [http://www.pasteur.fr/mlst/], Pasteur Institute,
- Harmsen D, Claus H, Witte W, Rothgänger J, Claus H, Turnwald D, Vogel U:Typing of methicillin-resistantStaphylococcus aureusin a university hospital setting by using novel software for spa repeat determination and database management. J Clin Microbiol. 2003, 41 (12): 5442-5448. 10.1128/JCM.41.12.5442-5448.2003.View ArticleGoogle Scholar
- Boucher HW, Corey GR:Epidemiology of methicillin-resistant Staphylococcus aureus. Clin Infect Dis. 2008, 46 (Suppl 5): 344-349.View ArticleGoogle Scholar
- Lindstedt B-A:Multiple-locus variable number tandem repeats analysis for genetic fingerprinting of pathogenic bacteria. Electrophoresis. 2005, 26 (13): 2567-2582. 10.1002/elps.200500096.View ArticleGoogle Scholar
- Jolley KA, Maiden MCJ:BIGSdb: Scalable analysis of bacterial genome variation at the population level. BMC Bioinformatics. 2010, 11: 595-10.1186/1471-2105-11-595.View ArticleGoogle Scholar
- Jolley KA, Bliss CM, Bennett JS, Bratcher HB, Brehony C, Colles FM, Wimalarathna H, Harrison OB, Sheppard SK, Cody AJ, Maiden MCJ:Ribosomal multilocus sequence typing: universal characterization of bacteria from domain to strain. Microbiology. 2012, 158 (Pt 4): 1005-1015.View ArticleGoogle Scholar
- Croucher NJ, Harris SR, Fraser C, Quail MA, Burton J, van der Linden M, McGee L, von Gottberg A, Song J-H, Ko KS, Pichon B, Baker S, Parry CM, Lambertsen LM, Shahinas D, Pillai DR, Mitchell TJ, Dougan G, Tomasz A, Klugman KP, Parkhill J, Hanage WP, Bentley SD:Rapid pneumococcal evolution in response to clinical interventions. Science. 2011, 331 (6016): 430-434. 10.1126/science.1198545.View ArticleGoogle Scholar
- Harris SR, Feil EJ, Quail MA, Nickerson EK, Chantratita N, Gardete S, Tavares A, Day N, Lindsay JA, Edgeworth JD, de Lencastre H, Parkhill J, Peacock SJ, Bentley SD, Holden MT G:Evolution of MRSA during hospital transmission and intercontinental spread. Science. 2010, 327 (5964): 469-474. 10.1126/science.1182395.View ArticleGoogle Scholar
- Bittner T, Smith B:Normalizing medical ontologies using basic formal ontology. Kooperative Versorgung, Vernetzte Forschung, Ubiquitäre Information (Proceedings of GMDS Innsbruck, 26–30 September 2004). 2004, Niebüll: Videel OHG, 199-201.Google Scholar
- Eilbeck K, Lewis SE, Mungall CJ, Yandell M, Stein L, Durbin R, Ashburner M:The sequence ontology: a tool for the unification of genome annotations. Genome Biol. 2005, 6 (5): 44-10.1186/gb-2005-6-5-r44.View ArticleGoogle Scholar
- Buttigieg PL, Morrison N, Smith B, Mungall CJ, Lewis SE, the ENVO Consortium:The environment ontology: contextualising biological and biomedical entities. J Biomed Semant. 2013, 4: 43-10.1186/2041-1480-4-43.View ArticleGoogle Scholar
- Peters B, the OBI Consortium:Ontology for biomedical investigations. International Conference on Biomedical Ontology (ICBO 2009). 2009. Available from Nature Precedings [http://dx.doi.org/10.1038/npre.2009.3623.1],
- Uniprot Core Ontology. [http://purl.uniprot.org/core/],
- The friend of a friend(foaf) project. [http://www.foaf-project.org/],
- The DBpedia Ontology. [http://wiki.dbpedia.org/Ontology],
- Redaschi N, UniProt Consortium:UniProt in RDF: Tackling data integration and distributed annotation with the semantic web. 3rd International Biocuration Conference. 2009. Available from Nature Precedings [http://dx.doi.org/10.1038/npre.2009.3193.1],
- Brinkman RR, Courtot M, Derom D, Fostel JM, He Y, Lord P, Malone J, Parkinson H, Peters B, Rocca-Serra P, Ruttenberg A, Sansone S-A, Soldatova LN, Stoeckert CJ, Turner JA, Zheng J, the OBI Consortium:Modeling biomedical experimental processes with OBI. J Biomed Semant. 2010, 1 (Suppl 1): 7-10.1186/2041-1480-1-S1-S7.View ArticleGoogle Scholar
- Bizer C, Lehmann J, Kobilarov G, Auer S, Becker C, Cyganiak R, Hellmann S:DBpedia – a crystallization point for the web of data. Web Semant Sci Serv Agents World Wide Web. 2009, 7 (3): 154-165. 10.1016/j.websem.2009.07.002.View ArticleGoogle Scholar
- Web Access Control. [http://www.w3.org/wiki/WebAccessControl],
- Dumbill E:Support online communities with foaf: How the friend-of-a-friend vocabulary addresses issues of accountability and privacy. 2002. IBM’s XML Watch, [http://web.archive.org/web/20070226182529/http://www-128.ibm.com/developerworks/xml/library/x-foaf2.html],
- Dumbill E:Tracking provenance of rdf data. 2003. IBM’s XML Watch, [http://web.archive.org/web/20050428083208/http://www-106.ibm.com/developerworks/xml/library/x-rdfprov.html],
- Xiang Z, Courtot M, Brinkman RR, Ruttenberg A, He Y:Ontofox: web-based support for ontology reuse. BMC Res notes. 2010, 3 (1): 175-10.1186/1756-0500-3-175.View ArticleGoogle Scholar
- Beckett D, Berners-Lee T, Prud’hommeaux E, Carothers G:Turtle: Terse RDF Triple Language. 2012. [http://www.w3.org/TR/2012/WD-turtle-20120710/],
- Bizer C, Seaborne A:D2RQ-treating non-RDF databases as virtual RDF graphs. Proceedings of the 3rd International Semantic Web Conference (ISWC2004). 2004, Hiroshima: Springer,Google Scholar
- Bizer C, Cyganiak R, Garbers G, Maresch O, Becker C:The D2RQ Platform v0. 7-Treating Non-RDF Relational Databases as Virtual RDF Graphs. 2009. User Manual and Language Specification [http://wifo5-03.informatik.uni-mannheim.de/bizer/d2rq/spec/20090810/],
- Maiden MCJ, van Rensburg MJJ, Bray JE, Earle SG, Ford SA, Jolley KA, McCarthy ND:MLST revisited: the gene-by-gene approach to bacterial genomics. Nat Rev Microbiol. 2013, 11 (10): 728-736. 10.1038/nrmicro3093.View ArticleGoogle Scholar
This article is published under license to BioMed Central Ltd. This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/4.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly credited. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated.