www.nature.com/scientificdata
OPEN
SUBJECT CATEGORIES
Research data
Publication
characteristics
Received: 10 December 2015
Accepted: 12 February 2016 Published: 15 March 2016
Comment: The FAIR Guiding Principles for scientic data management and stewardship
Mark D. Wilkinson et al.#
There is an urgent need to improve the infrastructure supporting the reuse of scholarly data. A diverse set of stakeholdersrepresenting academia, industry, funding agencies, and scholarly publishershave come together to design and jointly endorse a concise and measureable set of principles that we refer to as the FAIR Data Principles. The intent is that these may act as a guideline for those wishing to enhance the reusability of their data holdings. Distinct from peer initiatives that focus on the human scholar, the FAIR Principles put specic emphasis on enhancing the ability of machines to automatically nd and use the data, in addition to supporting its reuse by individuals. This Comment is the rst formal publication of the FAIR Principles, and includes the rationale behind them, and some exemplar implementations in the community.
Supporting discovery through good data management
Good data management is not a goal in itself, but rather is the key conduit leading to knowledge discovery and innovation, and to subsequent data and knowledge integration and reuse by the community after the data publication process. Unfortunately, the existing digital ecosystem surrounding scholarly data publication prevents us from extracting maximum benet from our research investments (e.g., ref. 1). Partially in response to this, science funders, publishers and governmental agencies are beginning to require data management and stewardship plans for data generated in publicly funded experiments. Beyond proper collection, annotation, and archival, data stewardship includes the notion of long-term care of valuable digital assets, with the goal that they should be discovered and re-used for downstream investigations, either alone, or in combination with newly generated data. The outcomes from good data management and stewardship, therefore, are high quality digital publications that facilitate and simplify this ongoing process of discovery, evaluation, and reuse in downstream studies. What constitutes good data management is, however, largely undened, and is generally left as a decision for the data or repository owner. Therefore, bringing some clarity around the goals and desiderata of good data management and stewardship, and dening simple guideposts to inform those who publish and/or preserve scholarly data, would be of great utility.
This article describes four foundational principlesFindability, Accessibility, Interoperability, and Reusabilitythat serve to guide data producers and publishers as they navigate around these obstacles, thereby helping to maximize the added-value gained by contemporary, formal scholarly digital publishing. Importantly, it is our intent that the principles apply not only to data in the conventional sense, but also to the algorithms, tools, and workows that led to that data. All scholarly digital research objects2from data to analytical pipelinesbenet from application of these principles, since all components of the research process must be available to ensure transparency, reproducibility, and reusability.
There are numerous and diverse stakeholders who stand to benet from overcoming these obstacles: researchers wanting to share, get credit, and reuse each others data and interpretations; professional data publishers offering their services; software and tool-builders providing data analysis and processing services such as reusable workows; funding agencies (private and public) increasingly
Correspondence and requests for materials should be addressed to B.M. (email: mailto:[email protected]
Web End [email protected] ). #A full list of authors and their afliations appears at the end of the paper.
SCIENTIFIC DATA | 3:160018 | DOI: 10.1038/sdata.2016.18 1
www.nature.com/sdata/
Box 1 | Terms and Abbreviations
BD2KBig Data 2 Knowledge, is a trans-NIH initiative established to enable biomedical research as a digital research enterprise, to facilitate discovery and support new knowledge, and to maximise community engagement.
DOIDigital Object Identier; a code used to permanently and stably identify (usually digital) objects. DOIs provide a standard mechanism for retrieval of metadata about the object, and generally a means to access the data object itself.
FAIRFindable, Accessible, Interoperable, Reusable.
FORCE11The Future of Research Communications and e-Scholarship; a community of scholars, librarians, archivists, publishers and research funders that has arisen organically to help facilitate the change toward improved knowledge creation and sharing, initiated in 2011.
Interoperabilitythe ability of data or tools from non-cooperating resources to integrate or work together with minimal effort.
JDDCPJoint Declaration of Data Citation Principles; Acknowledging data as a rst-class research output, and to support good research practices around data re-use, JDDCP proposes a set of guiding principles for citation of data within scholarly literature, another dataset, or any other research object.
RDFResource Description Framework; a globally-accepted framework for data and knowledge representation that is intended to be read and interpreted by machines.
concerned with long-term data stewardship; and a data science community mining, integrating and analysing new and existing data to advance discovery. To facilitate the reading of this manuscript by these diverse stakeholders, we provide denitions for common abbreviations in Box 1. Humans, however, are not the only critical stakeholders in the milieu of scientic data. Similar problems are encountered by the applications and computational agents that we task to undertake data retrieval and analysis on our behalf. These computational stakeholders are increasingly relevant, and demand as much, or more, attention as their importance grows. One of the grand challenges of data-intensive science, therefore, is to improve knowledge discovery through assisting both humans, and their computational agents, in the discovery of, access to, and integration and analysis of, task-appropriate scientic data and other scholarly digital objects.
For certain types of important digital objects, there are well-curated, deeply-integrated, special-purpose repositories such as Genbank3, Worldwide Protein Data Bank (wwPDB4), and UniProt5 in the life sciences; Space Physics Data Facility (SPDF; http://spdf.gsfc.nasa.gov/
Web End =http://spdf.gsfc.nasa.gov/ ) and Set of Identications, Measurements and Bibliography for Astronomical Data (SIMBAD6) in the space sciences. These foundational and critical core resources are continuously curating and capturing high-value reference datasets and ne-tuning them to enhance scholarly output, provide support for both human and mechanical users, and provide extensive tooling to access their content in rich, dynamic ways. However, not all datasets or even data types can be captured by, or submitted to, these repositories. Many important datasets emerging from traditional, low-throughput bench science dont t in the data models of these special-purpose repositories, yet these datasets are no less important with respect to integrative research, reproducibility, and reuse in general. Apparently in response to this, we see the emergence of numerous general-purpose data repositories, at scales ranging from institutional (for example, a single university), to open globally-scoped repositories such as Dataverse7,
FigShare (http://figshare.com
Web End =http://gshare.com ), Dryad8, Mendeley Data (https://data.mendeley.com/
Web End =https://data.mendeley.com/), Zenodo (http://zenodo.org/
Web End =http:// http://zenodo.org/
Web End =zenodo.org/ ), DataHub (http://datahub.io
Web End =http://datahub.io), DANS (http://www.dans.knaw.nl/
Web End =http://www.dans.knaw.nl/), and EUDat9. Such repositories accept a wide range of data types in a wide variety of formats, generally do not attempt to integrate or harmonize the deposited data, and place few restrictions (or requirements) on the descriptors of the data deposition. The resulting data ecosystem, therefore, appears to be moving away from centralization, is becoming more diverse, and less integrated, thereby exacerbating the discovery and re-usability problem for both human and computational stakeholders.
A specic example of these obstacles could be imagined in the domain of gene regulation and expression analysis. Suppose a researcher has generated a dataset of differentially-selected polyadenylation sites in a non-model pathogenic organism grown under a variety of environmental conditions that stimulate its pathogenic state. The researcher is interested in comparing the alternatively-polyadenylated genes in this local dataset, to other examples of alternative-polyadenylation, and the expression levels of these genesboth in this organism and related model organismsduring the infection process. Given that there is no special-purpose archive for differential polyadenylation data, and no model organism database for this pathogen, where does the researcher begin?
We will consider the current approach to this problem from a variety of data discovery and integration perspectives. If the desired datasets existed, where might they have been published, and how would one begin to search for them, using what search tools? The desired search would need to lter based on specic species, specic tissues, specic types of data (Poly-A, microarray, NGS), specic conditions (infection), and specic genesis that information (metadata) captured by the repositories, and if so, what formats is it in, is it searchable, and how? Once the data is discovered, can it be downloaded? In what format(s)? Can that format be easily integrated with private in-house data (the local dataset of alternative polyadenylation sites) as well as other data publications from third-parties and with the communitys core gene/protein data repositories? Can this integration be
SCIENTIFIC DATA | 3:160018 | DOI: 10.1038/sdata.2016.18 2
www.nature.com/sdata/
done automatically to save time and avoid copy/paste errors? Does the researcher have permission to use the data from these third-party researchers, under what license conditions, and who should be cited if a data-point is re-used?
Questions such as these highlight some of the barriers to data discovery and reuse, not only for humans, but even more so for machines; yet it is precisely these kinds of deeply and broadly integrative analyses that constitute the bulk of contemporary e-Science. The reason that we often need several weeks (or months) of specialist technical effort to gather the data necessary to answer such research questions is not the lack of appropriate technology; the reason is, that we do not pay our valuable digital objects the careful attention they deserve when we create and preserve them. Overcoming these barriers, therefore, necessitates that all stakeholdersincluding researchers, special-purpose, and general-purpose repositoriesevolve to meet the emergent challenges described above. The goal is for scholarly digital objects of all kinds to become rst class citizens in the scientic publication ecosystem, where the quality of the publicationand more importantly, the impact of the publicationis a function of its ability to be accurately and appropriately found, reused, and cited over time, by all stakeholders, both human and mechanical.
With this goal in-mind, a workshop was held in Leiden, Netherlands, in 2014, named Jointly Designing a Data Fairport. This workshop brought together a wide group of academic and private stakeholders all of whom had an interest in overcoming data discovery and reuse obstacles. From the deliberations at the workshop the notion emerged that, through the denition of, and widespread support for, a minimal set of community-agreed guiding principles and practices, all stakeholders could more easily discover, access, appropriately integrate and re-use, and adequately cite, the vast quantities of information being generated by contemporary data-intensive science. The meeting concluded with a draft formulation of a set of foundational principles that were subsequently elaborated in greater detailnamely, that all research objects should be Findable, Accessible, Interoperable and Reusable (FAIR) both for machines and for people. These are now referred to as the FAIR Guiding Principles. Subsequently, a dedicated FAIR working group, established by several members of the FORCE11 community10 ne-tuned and improved the Principles. The results of these efforts are reported here.
The signicance of machines in data-rich research environments
The emphasis placed on FAIRness being applied to both human-driven and machine-driven activities, is a specic focus of the FAIR Guiding Principles that distinguishes them from many peer initiatives (discussed in the subsequent section). Humans and machines often face distinct barriers when attempting to nd and process data on the Web. Humans have an intuitive sense of semantics (the meaning or intent of a digital object) because we are capable of identifying and interpreting a wide variety of contextual cues, whether those take the form of structural/visual/iconic cues in the layout of a Web page, or the content of narrative notes. As such, we are less likely to make errors in the selection of appropriate data or other digital objects, although humans will face similar difculties if sufcient contextual metadata is lacking. The primary limitation of humans, however, is that we are unable to operate at the scope, scale, and speed necessitated by the scale of contemporary scientic data and complexity of e-Science. It is for this reason that humans increasingly rely on computational agents to undertake discovery and integration tasks on their behalf. This necessitates machines to be capable of autonomously and appropriately acting when faced with the wide range of types, formats, and access-mechanisms/protocols that will be encountered during their self-guided exploration of the global data ecosystem. It also necessitates that the machines keep an exquisite record of provenance such that the data they are collecting can be accurately and adequately cited. Assisting these agents, therefore, is a critical consideration for all participants in the data management and stewardship processfrom researchers and data producers to data repository hosts.
Throughout this paper, we use the phrase machine actionable to indicate a continuum of possible states wherein a digital object provides increasingly more detailed information to an autonomously-acting, computational data explorer. This information enables the agentto a degree dependent on the amount of detail providedto have the capacity, when faced with a digital object never encountered before, to: a) identify the type of object (with respect to both structure and intent), b) determine if it is useful within the context of the agents current task by interrogating metadata and/ or data elements, c) determine if it is usable, with respect to license, consent, or other accessibility or use constraints, and d) take appropriate action, in much the same manner that a human would.
For example, a machine may be capable of determining the data-type of a discovered digital object, but not capable of parsing it due to it being in an unknown format; or it may be capable of processing the contained data, but not capable of determining the licensing requirements related to the retrieval and/or use of that data. The optimal statewhere machines fully understand and can autonomously and correctly operate-on a digital objectmay rarely be achieved. Nevertheless, the FAIR principles provide steps along a path toward machine-actionability; adopting, in whole or in part, the FAIR
SCIENTIFIC DATA | 3:160018 | DOI: 10.1038/sdata.2016.18 3
www.nature.com/sdata/
Box 2 | The FAIR Guiding Principles
To be Findable:
F1. (meta)data are assigned a globally unique and persistent identierF2. data are described with rich metadata (dened by R1 below)
F3. metadata clearly and explicitly include the identier of the data it describes F4. (meta)data are registered or indexed in a searchable resource
To be Accessible:
A1. (meta)data are retrievable by their identier using a standardized communications protocol A1.1 the protocol is open, free, and universally implementableA1.2 the protocol allows for an authentication and authorization procedure, where necessary A2. metadata are accessible, even when the data are no longer available
To be Interoperable:
I1. (meta)data use a formal, accessible, shared, and broadly applicable language for knowledge representation. I2. (meta)data use vocabularies that follow FAIR principlesI3. (meta)data include qualied references to other (meta)data
To be Reusable:
R1. meta(data) are richly described with a plurality of accurate and relevant attributes R1.1. (meta)data are released with a clear and accessible data usage licenseR1.2. (meta)data are associated with detailed provenanceR1.3. (meta)data meet domain-relevant community standards
principles, leads the resource along the continuum towards this optimal state. In addition, the idea of being machine-actionable applies in two contextsrst, when referring to the contextual metadata surrounding a digital object (what is it?), and second, when referring to the content of the digital object itself (how do I process it/integrate it?). Either, or both of these may be machine-actionable, and each forms its own continuum of actionability.
Finally, we wish to draw a distinction between data that is machine-actionable as a result of specic investment in software supporting that data-type, for example, bespoke parsers that understand life science wwPDB les or space science Space Physics Archive Search and Extract (SPASE) les, and data that is machine-actionable exclusively through the utilization of general-purpose, open technologies. To reiterate the earlier pointultimate machine-actionability occurs when a machine can make a useful decision regarding data that it has not encountered before. This distinction is important when considering both (a) the rapidly growing and evolving data environment, with new technologies and new, more complex data-types continuously being developed, and (b) the growth of general-purpose repositories, where the data-types likely to be encountered by an agent are unpredictable. Creating bespoke parsers, in all computer languages, for all data-types and all analytical tools that require those data-types, is not a sustainable activity. As such, the focus on assisting machines in their discovery and exploration of data through application of more generalized interoperability technologies and standards at the data/repository level, becomes a rst-priority for good data stewardship.
The FAIR Guiding Principles in detail
Representatives of the interested stakeholder-groups, discussed above, coalesced around four core desideratathe FAIR Guiding Principlesand limited elaboration of these, which have been rened (Box 2) from the meetings original draft, available at (https://www.force11.org/node/6062
Web End =https://www.force11.org/node/6062 ). A separate document that dynamically addresses community discussion relating to clarications and explanations of the principles, and detailed guidelines for and examples of FAIR implementations, is currently being constructed (http://datafairport.org/fair-principles-living-document-menu
Web End =http://datafairport.org/fair-principles-living-document-menu). The FAIR Guiding Principles describe distinct considerations for contemporary data publishing environments with respect to supporting both manual and automated deposition, exploration, sharing, and reuse. While there have been a number of recent, often domain-focused publications advocating for specic improvements in practices relating to data management and archival1,11,12, FAIR differs in that it describes concise, domain-independent, high-level principles that can be applied to a wide range of scholarly outputs. Throughout the Principles, we use the phrase (meta)data in cases where the Principle should be applied to both metadata and data.
The elements of the FAIR Principles are related, but independent and separable. The Principles dene characteristics that contemporary data resources, tools, vocabularies and infrastructures should exhibit to assist discovery and reuse by third-parties. By minimally dening each guiding principle, the barrier-to-entry for data producers, publishers and stewards who wish to make their data holdings FAIR is purposely maintained as low as possible. The Principles may be adhered to in any combination and incrementally, as data providers publishing environments evolve to increasing degrees of FAIRness. Moreover, the modularity of the Principles, and their distinction between data and metadata, explicitly support a wide range of special circumstances. One such example is highly sensitive or personally-identiable data, where publication of rich metadata to facilitate discovery, including clear rules regarding the process for accessing the data, provides a high degree of FAIRness even in the absence of FAIR publication of the data itself. A second example involves the publication
SCIENTIFIC DATA | 3:160018 | DOI: 10.1038/sdata.2016.18 4
www.nature.com/sdata/
of non-data research objects. Analytical workows, for example, are a critical component of the scholarly ecosystem, and their formal publication is necessary to achieve both transparency and scientic reproducibility. The FAIR principles can equally be applied to these non-data assets, which need to be identied, described, discovered, and reused in much the same manner as data.
Specic exemplar efforts that provide varying levels of FAIRness are detailed later in this document. Additional issues, however, remain to be addressed. First, when community-endorsed vocabularies or other (meta)data standards do not include the attributes necessary to achieve rich annotation, there are two possible solutions: either publish an extension of an existing, closely related vocabulary, orin the extreme casecreate and explicitly publish a new vocabulary resource, following FAIR principles (I2). Second, to explicitly identify the standard chosen when more than one vocabulary or other (meta)data standard is available, and given that for instance in the life sciences there are over 600 content standards, the BioSharing registry (https://biosharing.org/
Web End =https://biosharing.org/) can be of use as it describes the standards in detail, including versions where applicable.
The Principles precede implementation
These high-level FAIR Guiding Principles precede implementation choices, and do not suggest any specic technology, standard, or implementation-solution; moreover, the Principles are not, themselves, a standard or a specication. They act as a guide to data publishers and stewards to assist them in evaluating whether their particular implementation choices are rendering their digital research artefacts Findable, Accessible, Interoperable, and Reusable. We anticipate that these high level principles will enable a broad range of integrative and exploratory behaviours, based on a wide range of technology choices and implementations. Indeed, many repositories are already implementing various aspects of FAIR using a variety of technology choices and several examples are detailed in the next section; examples include Scientic Data itself and how narrative data articles are anchored to a progressively FAIR structured metadata.
Examples of FAIRness, and the resulting value-added
Dataverse7: Dataverse is an open-source data repository software installed in dozens of institutions globally to support public community repositories or institutional research data repositories. Harvard Dataverse, with more than 60,000 datasets, is the largest of the current Dataverse repositories, and is open to all researchers from all research elds. Dataverse generates a formal citation for each deposit, following the standard dened by Altman and King13. Dataverse makes the Digital Object Identier (DOI), or other persistent identiers (Handles), public when the dataset is published (F). This resolves to a landing page, providing access to metadata, data les, dataset terms, waivers or licenses, and version information, all of which is indexed and searchable (F, A, and R). Deposits include metadata, data les, and any complementary les (such as documentation or code) needed to understand the data and analysis (R). Metadata is always public, even if the data are restricted or removed for privacy issues (F, A). This metadata is offered at three levels, extensively supporting the I and R FAIR principles: 1) data citation metadata, which maps to DataCite schema or Dublin Core Terms, 2) domain-specic metadata, which when possible maps to metadata standards used within a scientic domain, and 3) le-level metadata, which can be deep and extensive for tabular data les (including column-level metadata). Finally, Dataverse provides public machine-accessible interfaces to search the data, access the metadata and download the data les, using a token to grant access when data les are restricted (A).
FAIRDOM (http://fair-dom.org/about
Web End =http://fair-dom.org/about): integrates the SEEK14 and openBIS15 platforms to produce a FAIR data and model management facility for Systems Biology. Individual research assets (or aggregates of data and models) are identied with unique and persistent HTTP URLs, which can be registered with DOIs for publication (F). Assets can be accessed over the Web in a variety of formats appropriate for individuals and/or their computers (RDF, XML) (I). Research assets are annotated with rich metadata, using community standards, formats and ontologies (I). The metadata is stored as RDF to enable interoperability and assets can be downloaded for reuse (R).
ISA16: is a community-driven metadata tracking framework to facilitate standards-compliant collection, curation, management and reuse of life science datasets. ISA provides progressively FAIR structured metadata to Nature Scientic Datas Data Descriptor articles, and many GigaScience data papers, and underpins the EBI MetaboLights database among other data resources. At the heart is a general-purpose, extensible ISA model, originally only available as a tabular representation but subsequently enhanced as an RDF-based representation17, and JSON serializations to enable the I and R, becoming FAIR when published as linked data (http://elixir-uk.org/node-events/201cisa-as-a-fair-research-object201d-hack-the-spec-event-1
Web End =http://elixir-uk.org/node-events/201cisa-as-a- http://elixir-uk.org/node-events/201cisa-as-a-fair-research-object201d-hack-the-spec-event-1
Web End =fair-research-object201d-hack-the-spec-event-1 ) and complementing other research objects18.
Open PHACTS19: Open PHACTS is a data integration platform for information pertaining to drug discovery. Access to the platform is mediated through a machine-accessible interface20 which provides multiple representations that are both human (HTML) and machine readable (RDF, JSON,
SCIENTIFIC DATA | 3:160018 | DOI: 10.1038/sdata.2016.18 5
www.nature.com/sdata/
Box 3 | Emergent community/collaborative initiatives with FAIR as a core focus or activity
bioCADDIE (https://biocaddie.org
Web End =https://biocaddie.org): The NIH BD2K biomedical and healthCAre Data Discovery Index Ecosystem (bioCADDIE) consortium works to develop a Data Discovery Index (DDI) prototype, which is set to be as transformative and impactful for data as PubMed for the biomedical literature . The DDI focuses on nding (F) and accessing (A) the datasets stored across different sources, and progressively works to identify relevant metadata (I) and maps them to community standards (R), linking to BioSharing.
CEDAR : The Center for Expanded Data Annotation and Retrieval (CEDAR) is an NIH BD2K funded center of excellence to develop tools and technologies that reduce the burden of authoring and enhancing metadata that meet community-based standards. CEDAR will enable the creation of metadata templates that implement community based standards for experimental metadata, from BioSharing (https://biosharing.org
Web End =https://biosharing.org), and that will be uniquely identiable and retrievable with HTTP URIs, and annotated with vocabularies and ontologies drawn from BioPortal (http://bioportal.bioontology.org
Web End =http://bioportal.bioontology.org) (F,A,I,R). These templates will guide users to create rich metadata with unique and stable HTTP identiers (F) that can be retrieved using HTTP (A) and accessible in a variety of formats (JSON-LD, TURTLE, RDF/XML, CSV, etc) (I). These metadata will use community standards, as dened by the template, and include provenance and data usage (R).
These two projects, among others, provide tools and or collaborative opportunities for those who wish to improve the FAIRness of their data.
XML, CSV, etc), providing the A facet of FAIRness. The interface allows multiple URLs to be used to access information about a particular entity through a mappings service (F and A). Thus, a user can provide a ChEMBL URL to retrieve information sourced from, for example, Chemspider or DrugBank. Each call provides a canonical URL in its response (A and I). All data sources used are described using standardized dataset descriptions, following the global VoID standard, with rich provenance (R and I). All interface features are described using RDF following the Linked Data API specication (A). Finally, a majority of the datasets are described using community agreed upon ontologies (I).
wwPDB4,21: wwPDB is a special-purpose, intensively-curated data archive that hosts information about experimentally-determined 3D structures of proteins and nucleic acids. All wwPDB entries are stably hosted on an FTP server (A) and represented in machine-readable formats (text and XML); the latter are machine-actionable using the metadata provided by the wwPDB conforming to the Macromolecular Information Framework (mmCIF22), a data standard of the International Union of Crystallography (IUCr) (F,I for humans, F,I for IUCr-aware machines). The wwPDB metadata contains cross-references to common identiers such as PubMed and NCBI Taxonomy, and their wwPDB metadata are described in data dictionaries and schema documents (http://mmcif.wwpdb.org
Web End =http://mmcif.wwpdb.org and http://pdbml.wwpdb.org
Web End =http://pdbml.wwpdb.org ) which conform to the IUCr data standard for the chemical and structural biology domains (R). A variety of software tools are available to interpret both wwPDB data and meta-data (I,R for humans, I,R for machines with this software). Each entry is represented by a DOI (F, A for humans and machines). The DOI resolves to a zipped le which requires special software for further interrogation/interpretation. Other wwPDB access points2325
provide access to wwPDB records through URLs that are likely to be stable in the long-term (F), and all data and metadata is searchable through one or more of the wwPDB-afliated websites (F)
UniProt26: UniProt is a comprehensive resource for protein sequence and annotation data. All entries are uniquely identied by a stable URL, that provides access to the record in a variety of formats including a web page, plain-text, and RDF (F and A). The record contains rich metadata (F) that is both human-readable (HTML) and machine-readable (text and RDF), where the RDF formatted response utilizes shared vocabularies and ontologies such as UniProt Core, FALDO, and ECO (I). Interlinking with more than 150 different databases, every UniProt record has extensive links into, for example, PubMed, enabling rich citation. These links are machine-actionable in the RDF representation (R). Finally, in the RDF representation, the UniProt Core Ontology explicitly types all records, leaving no ambiguityneither for humans nor machinesabout what the data represents (R), enabling fully-automated retrieval of records and cross-referencing information.
In addition to, and in support of, communities and resources that are already pursuing FAIR objectives, the Data Citation Implementation Group of Force11 has published specic technical recommendations for how to implement many of the principles27, with a particular focus on identiers and their resolution, persistence, and metadata accessibility especially related to citation. In addition, the Skunkworks group that emerged from the Lorentz Workshop has been creating software supporting infrastructures28 that are, end-to-end, compatible with FAIR principles, and can be implemented over existing repositories. These code modules have a particular focus on metadata publication and searchability, compatibility in cases of strict privacy considerations, and the extremely difcult problem of data and metadata interoperability (manuscript in preparation). Finally, there are several emergent projects, some listed in Box 3, for which FAIR is a key objective. These projects may provide valuable advice and guidance for those wishing to become more FAIR.
FAIRness is a prerequisite for proper data management and data stewardshipThe ideas within the FAIR Guiding Principles reect, combine, build upon and extend previous work by both the Concept Web Alliance (https://conceptweblog.wordpress.com/
Web End =https://conceptweblog.wordpress.com/) partners, who focused on machine-actionability and harmonization of data structures and semantics, and by the scientic and scholarly organizations that developed the Joint Declaration of Data Citation Principles (JDDCP29),
SCIENTIFIC DATA | 3:160018 | DOI: 10.1038/sdata.2016.18 6
www.nature.com/sdata/
who focused on primary scholarly data being made citable, discoverable and available for reuse, so as to be capable of supporting more rigorous scholarship. An attempt to dene the similarities and overlaps between the FAIR Principles and the JDDCP is provided at (https://www.force11.org/node/6062
Web End =https://www.force11.org/node/ https://www.force11.org/node/6062
Web End =6062 ). The FAIR Principles are also complementary to the Data Seal of Approval (DSA) (http://datasealofapproval.org/media/filer_public/2013/09/27/guidelines_2014-2015.pdf
Web End =http:// http://datasealofapproval.org/media/filer_public/2013/09/27/guidelines_2014-2015.pdf
Web End =datasealofapproval.org/media/ler_public/2013/09/27/guidelines_2014-2015.pdf ) in that they share the general aim to render data re-usable for users other than those who originally generated them. While the DSA focuses primarily on the responsibilities and conduct of data producers and repositories, FAIR focuses primarily on the data itself. Clearly, the broader community of stakeholders is coalescing around a set of common, dovetailed visions spanning all facets of the scholarly data publishing ecosystem.
The end result, when implemented, will be more rigorous management and stewardship of these valuable digital resources, to the benet of the entire academic community. As stated at the outset, good data management and stewardship is not a goal in itself, but rather a pre-condition supporting knowledge discovery and innovation. Contemporary e-Science requires data to be Findable, Accessible, Interoperable, and Reusable in the long-term, and these objectives are rapidly becoming expectations of agencies and publishers. We demonstrate, therefore, that the FAIR Data Principles provide a set of mileposts for data producers and publishers. They guide the implementation of the most basic levels of good Data Management and Stewardship practice, thus helping researchers adhere to the expectations and requirements of their funding agencies. We call on all data producers and publishers to examine and implement these principles, and actively participate with the FAIR initiative by joining the Force11 working group. By working together towards shared, common goals, the valuable data produced by our community will gradually achieve the critical goals of FAIRness.
References
1. Roche, D. G., Kruuk, L. E. B., Lanfear, R. & Binning, S. A. Public Data Archiving in Ecology and Evolution: How Well Are We Doing? PLOS Biol. 13, e1002295 (2015).
2. Bechhofer, S. et al. Research Objects: Towards Exchange and Reuse of Digital Knowledge. Nat. Preced. doi:http://dx.doi.org/10.1038/npre.2010.4626.1
Web End =10.1038/ http://dx.doi.org/10.1038/npre.2010.4626.1
Web End =npre.2010.4626.1 (2010).
3. Benson, D. A. et al. GenBank. Nucleic Acids Res. 41, D36D42 (2013).4. Berman, H., Henrick, K. & Nakamura, H. Announcing the worldwide Protein Data Bank. Nat. Struct. Biol. 10, 980980 (2003).5. The Uniprot Consortium. UniProt: a hub for protein information. Nucleic Acids Res. 43, D204D212 (2015).6. Wenger, M. et al. The SIMBAD astronomical database-The CDS reference database for astronomical objects. Astron. Astrophys. Suppl. Ser. 143, 922 (2000).
7. Crosas, M. "The Dataverse Network: An Open-Source Application for Sharing, Discovering and Preserving Data". D-Lib Mag 17(1), p2 (2011).8. White, H. C., Carrier, S., Thompson, A., Greenberg, J. & Scherle, R. The Dryad data repository: A Singapore framework metadata architecture in a DSpace environment. Univ. Gttingen, p157 (2008).
9. Lecarpentier, D. et al. EUDAT: A New Cross-Disciplinary Data Infrastructure for Science. Int. J. Digit. Curation 8, 279287 (2013).
10. Martone, M. E. FORCE11: Building the Future for Research Communications and e-Scholarship. Bioscience 65, 635 (2015).11. White, E. et al. Nine simple ways to make it easier to (re)use your data. Ideas Ecol. Evol. 6 (2013).12. Sandve, G. K., Nekrutenko, A., Taylor, J. & Hovig, E. Ten Simple Rules for Reproducible Computational Research. PLoS Comput. Biol. 9, e1003285 (2013).
13. Altman, M. & King, G. in D-Lib Magazine 13, no. 3/4 (2007).14. Wolstencroft, K. et al. SEEK: a systems biology data and model management platform. BMC Syst. Biol. 9, 33 (2015).15. Bauch, A. et al. openBIS: a exible framework for managing and analyzing complex data in biology research. BMC Bioinformatics 12, 468 (2011).
16. Sansone, S.-A. et al. Toward interoperable bioscience data. Nat. Genet. 44, 121126 (2012).17. Gonzlez-Beltrn, A., Maguire, E., Sansone, S.-A. & Rocca-Serra, P. linkedISA: semantic representation of ISA-Tab experimental metadata. BMC Bioinformatics 15, S4 (2014).
18. Gonzlez-Beltrn, A. et al. From Peer-Reviewed to Peer-Reproduced in Scholarly Publishing: The Complementary Roles of Data Models and Workows in Bioinformatics. PLoS ONE 10, e0127612 (2015).
19. Harland, L. Open PHACTS: A Semantic Knowledge Infrastructure for Public and Commercial Drug Discovery Research. Knowl. Eng. Knowl. Manag. Lect. Notes Comput. Sci. 7603/2012, 17 (2012).
20. Groth, P. et al. API-centric Linked Data integration: The Open PHACTS Discovery Platform case study. Web Semant. Sci. Serv. Agents World Wide Web 29, 1218 (2014).
21. Berman, H. M. et al. The Protein Data Bank. Nucleic Acids Res. 28, 235242 (2000).22. Bourne, P. E., Berman, H. M., Watenpaugh, K., Westbrook, J. D. & Fitzgerald, P. M. D. The macromolecular crystallographic information le (mmCIF). Meth. Enzym 277, 571590 (1997).
23. Rose, P. W. et al. The RCSB Protein Data Bank: views of structural biology for basic and applied research and education. Nucleic Acids Res. 43, D345D356 (2015).
24. Kinjo, A. R. et al. Protein Data Bank Japan (PDBj): maintaining a structural data archive and resource description framework format. Nucleic Acids Res. 40, D453D460 (2012).
25. Gutmanas, A. et al. PDBe: Protein Data Bank in Europe. Nucleic Acids Res. 42, D285D291 (2014).26. UniProt Consortium. UniProt: a hub for protein information. Nucleic Acids Res. 43, D204D212 (2015).27. Starr, J. et al. Achieving human and machine accessibility of cited data in scholarly publications. PeerJ Comput. Sci. 1, e1 (2015).28. Wilkinson, M., Dumontier, M. & Durbin, P. DataFairPort: The Perl libraries version 0.231 doi:http://dx.doi.org/10.5281/zenodo.33584
Web End =10.5281/zenodo.33584 (2015).29. Data Citation Synthesis Group: Joint Declaration of Data Citation Principles. San Diego CA: FORCE11. https://www.force11.org/datacitation
Web End =https://www.force11.org/ https://www.force11.org/datacitation
Web End =datacitation (2014).
30. Ohno-machado, L. et al. NIH BD2K bioCADDIE white paperData Discovery Index. http://dx.doi.org/10.6084/m9.figshare.1362572
Web End =http://dx.doi.org/10.6084/m9.g- http://dx.doi.org/10.6084/m9.figshare.1362572
Web End =share.1362572 (2015).
SCIENTIFIC DATA | 3:160018 | DOI: 10.1038/sdata.2016.18 7
www.nature.com/sdata/
31. NIH BD2K bioCADDIE WG3 Members. WG3-MetadataSpecications: NIH BD2K bioCADDIE Data Discovery Index WG3 Metadata Specication v1 doi:http://dx.doi.org/10.5281/zenodo.28019
Web End =10.5281/zenodo.28019 (2015).
32. Musen, M. A. et al. The center for expanded data annotation and retrieval. J. Am. Med. Informatics Assoc. 22, 11481152 (2015).
Acknowledgements
The original Lorentz Workshop Jointly Designing a Data FAIRport was organized by Barend Mons in collaboration with and co-sponsored by the Lorentz center, The Dutch Techcenter for the Life Sciences and the Netherlands eScience Center. The principles and themes described in this manuscript represent the signicant voluntary contributions and participation of the authors at, and/or subsequent to, this workshop and from the wider Force11, BD2K and ELIXIR communities. We also acknowledge and thank the organizers and backers of the NBDC/DBCLS BioHackathon 2015, where several of the authors made signicant revisions to the FAIR Principles.
Author Contributions
M.W. was the primary author of the manuscript, and participated extensively in the drafting and editing of the FAIR Principles. M.D. was signicantly involved in the drafting of the FAIR Principles. B.M. conceived of the FAIR Data Initiative, contributed extensively to the drafting of the principles, and to this manuscript text. All other authors are listed alphabetically, and contributed to the manuscript either by their participation in the initial workshop and/or by editing or commenting on the manuscript text.
Additional Information
Competing nancial interests: M.A. is the Nature Genetics Editor in Chief; S.A.S. is Scientic Datas Honorary Academic Editor and consultant.
How to cite this article: Wilkinson, M. D. et al. The FAIR Guiding Principles for scientic data management and stewardship. Sci. Data 3:160018 doi: 10.1038/sdata.2016.18 (2016).
This work is licensed under a Creative Commons Attribution 4.0 International License. The images or other third party material in this article are included in the articles Creative Commons license, unless indicated otherwise in the credit line; if the material is not included under the Creative Commons license, users will need to obtain permission from the license holder to reproduce the material. To view a copy of this license, visit http://creativecommons.org/licenses/by/4.0
Web End =http://creativecommons.org/licenses/by/4.0
Mark D. Wilkinson1, Michel Dumontier2, IJsbrand Jan Aalbersberg3, Gabrielle Appleton3, Myles Axton4, Arie Baak5, Niklas Blomberg6, Jan-Willem Boiten7,Luiz Bonino da Silva Santos8, Philip E. Bourne9, Jildau Bouwman10, Anthony J. Brookes11, Tim Clark12, Merc Crosas13, Ingrid Dillo14, Olivier Dumon3, Scott Edmunds15,Chris T. Evelo16, Richard Finkers17, Alejandra Gonzalez-Beltran18, Alasdair J.G. Gray19, Paul Groth3, Carole Goble20, Jeffrey S. Grethe21, Jaap Heringa22, Peter A.C. t Hoen23, Rob Hooft24, Tobias Kuhn25, Ruben Kok22, Joost Kok26, Scott J. Lusher27,Maryann E. Martone28, Albert Mons29, Abel L. Packer30, Bengt Persson31,Philippe Rocca-Serra18, Marco Roos32, Rene van Schaik33, Susanna-Assunta Sansone18, Erik Schultes34, Thierry Sengstag35, Ted Slater36, George Strawn37, Morris A. Swertz38, Mark Thompson32, Johan van der Lei39, Erik van Mulligen39, Jan Velterop40,Andra Waagmeester41, Peter Wittenburg42, Katherine Wolstencroft43, Jun Zhao44& Barend Mons45,46,47
1Center for Plant Biotechnology and Genomics, Universidad Politcnica de Madrid, Madrid 28223, Spain.
2Stanford University, Stanford 94305-5411, USA. 3Elsevier, Amsterdam 1043 NX, The Netherlands. 4Nature Genetics, New York 10004-1562, USA. 5Euretos and Phortos Consultants, Rotterdam 2741 CA, The Netherlands.
6ELIXIR, Wellcome Genome Campus, Hinxton CB10 1SA, UK. 7Lygature, Eindhoven 5656 AG, The Netherlands.
8Vrije Universiteit Amsterdam, Dutch Techcenter for Life Sciences, Amsterdam 1081 HV, The Netherlands.
9Ofce of the Director, National Institutes of Health, Rockville 20892, USA. 10TNO, Zeist 3700 AJ, The Netherlands. 11Department of Genetics, University of Leicester, Leicester LE1 7RH, UK. 12Harvard Medical School, Boston, Massachusetts MA 02115, USA. 13Harvard University, Cambridge, Massachusetts MA 02138, USA. 14Data Archiving and Networked Services (DANS), The Hague 2593 HW, The Netherlands. 15GigaScience, Beijing Genomics Institute, Shenzhen 518083, China. 16Department of Bioinformatics, Maastricht University, Maastricht 6200 MD, The Netherlands. 17Wageningen UR Plant Breeding, Wageningen 6708 PB, The Netherlands. 18Oxford e-Research Center, University of Oxford, Oxford OX1 3QG, UK. 19Heriot-Watt University, Edinburgh EH14 4AS, UK. 20School of Computer Science, University of Manchester, Manchester M13 9PL, UK.
21Center for Research in Biological Systems, School of Medicine, University of California San Diego, La Jolla, California 92093-0446, USA. 22Dutch Techcenter for the Life Sciences, Utrecht 3501 DE, The Netherlands.
23Department of Human Genetics, Leiden University Medical Center, Dutch Techcenter for the Life Sciences, Leiden 2300 RC, The Netherlands. 24Dutch TechCenter for Life Sciences and ELIXIR-NL, Utrecht 3501 DE, The
SCIENTIFIC DATA | 3:160018 | DOI: 10.1038/sdata.2016.18 8
www.nature.com/sdata/
Netherlands. 25VU University Amsterdam, Amsterdam 1081 HV, The Netherlands. 26Leiden Center of Data Science, Leiden University, Leiden 2300 RA, The Netherlands. 27Netherlands eScience Center, Amsterdam 1098 XG, The Netherlands. 28National Center for Microscopy and Imaging Research, UCSD, San Diego 92103, USA.
29Phortos Consultants, San Diego 92011, USA. 30SciELO/FAPESP Program, UNIFESP Foundation, So Paulo 05468-901, Brazil. 31Bioinformatics Infrastructure for Life Sciences (BILS), Science for Life Laboratory, Dept of Cell and Molecular Biology, Uppsala University, S-751 24, Uppsala, Sweden. 32Leiden University Medical Center, Leiden 2333 ZA, The Netherlands. 33Bayer CropScience, Gent Area 1831, Belgium. 34Leiden Institute for Advanced Computer Science, Leiden University Medical Center, Leiden 2300 RA, The Netherlands. 35Swiss Institute of Bioinformatics and University of Basel, Basel 4056, Switzerland. 36Cray, Inc., Seattle 98164, USA.
37Unafliated. 38University Medical Center Groningen (UMCG), University of Groningen, Groningen 9713 GZ, The Netherlands. 39Erasmus MC, Rotterdam 3015 CE, The Netherlands. 40Independent Open Access and Open Science Advocate, Guildford GU1 3PW, UK. 41Micelio, Antwerp 2180, Belgium. 42Max Planck Compute and Data Facility, MPS, Garching 85748, Germany. 43Leiden Institute of Advanced Computer Science, Leiden University, Leiden 2333 CA, The Netherlands. 44Department of Computer Science, Oxford University, Oxford OX1 3QD, UK.
45Leiden University Medical Center, Leiden and Dutch TechCenter for Life Sciences, Utrecht 2333 ZA, The Netherlands. 46Netherlands eScience Center, Amsterdam 1098 XG, The Netherlands. 47Erasmus MC, Rotterdam 3015 CE, The Netherlands.
SCIENTIFIC DATA | 3:160018 | DOI: 10.1038/sdata.2016.18 9
You have requested "on-the-fly" machine translation of selected content from our databases. This functionality is provided solely for your convenience and is in no way intended to replace human translation. Show full disclaimer
Neither ProQuest nor its licensors make any representations or warranties with respect to the translations. The translations are automatically generated "AS IS" and "AS AVAILABLE" and are not retained in our systems. PROQUEST AND ITS LICENSORS SPECIFICALLY DISCLAIM ANY AND ALL EXPRESS OR IMPLIED WARRANTIES, INCLUDING WITHOUT LIMITATION, ANY WARRANTIES FOR AVAILABILITY, ACCURACY, TIMELINESS, COMPLETENESS, NON-INFRINGMENT, MERCHANTABILITY OR FITNESS FOR A PARTICULAR PURPOSE. Your use of the translations is subject to all use restrictions contained in your Electronic Products License Agreement and by using the translation functionality you agree to forgo any and all claims against ProQuest or its licensors for your use of the translation functionality and any output derived there from. Hide full disclaimer
Copyright Nature Publishing Group Mar 2016
Abstract
There is an urgent need to improve the infrastructure supporting the reuse of scholarly data. A diverse set of stakeholders--representing academia, industry, funding agencies, and scholarly publishers--have come together to design and jointly endorse a concise and measureable set of principles that we refer to as the FAIR Data Principles. The intent is that these may act as a guideline for those wishing to enhance the reusability of their data holdings. Distinct from peer initiatives that focus on the human scholar, the FAIR Principles put specific emphasis on enhancing the ability of machines to automatically find and use the data, in addition to supporting its reuse by individuals. This Comment is the first formal publication of the FAIR Principles, and includes the rationale behind them, and some exemplar implementations in the community.
You have requested "on-the-fly" machine translation of selected content from our databases. This functionality is provided solely for your convenience and is in no way intended to replace human translation. Show full disclaimer
Neither ProQuest nor its licensors make any representations or warranties with respect to the translations. The translations are automatically generated "AS IS" and "AS AVAILABLE" and are not retained in our systems. PROQUEST AND ITS LICENSORS SPECIFICALLY DISCLAIM ANY AND ALL EXPRESS OR IMPLIED WARRANTIES, INCLUDING WITHOUT LIMITATION, ANY WARRANTIES FOR AVAILABILITY, ACCURACY, TIMELINESS, COMPLETENESS, NON-INFRINGMENT, MERCHANTABILITY OR FITNESS FOR A PARTICULAR PURPOSE. Your use of the translations is subject to all use restrictions contained in your Electronic Products License Agreement and by using the translation functionality you agree to forgo any and all claims against ProQuest or its licensors for your use of the translation functionality and any output derived there from. Hide full disclaimer