FAIR documentation Contentathon changes

_data/CONTRIBUTORS.yaml

@@ -369,4 +369,10 @@ Pablo Mata:
   email: pmata@externos.isciii.es
   orcid: 0009-0001-6887-0382
   role: editor
+  affiliation: ISCIII / ELIXIR-ES
+Enrique Sapena Ventura:
+  git: ESapenaVentura
+  email: enrique.sapena@isciii.es
+  orcid: 0000-0003-4554-9040
+  role: editor
   affiliation: ISCIII / ELIXIR-ES

_data/tool_and_resource_list.yml

@@ -1868,3 +1868,11 @@
   registry:
     biotools: comet
   url: https://github.com/damayanthiHerath/comet
+- description: AI/ML-powered tool to pseudonymize text/log files in order to meet Personally Identifiable Information (PII) regulation laws (e.g. European GDPR, US CCPA).
+  id: orion-pseudonymization-tool
+  name: Orion Pseudonymization Tool
+  url: https://www.orioninc.com/products/pseudonymization-tool/
+- description: AI-powered service to anonymize sensitive health data into high-quality, privacy-compliant assets.
+  id: veil-ai
+  name: Veil.AI
+  url: https://veil.ai/

data-description/fair-data.md

@@ -1,7 +1,7 @@
 ---
 title: FAIR data
 description: Apply the FAIR principles to infectious disease research data, emphasizing documentation, accessibility, metadata standards, and interoperability to promote transparency, collaboration, and reuse.
-contributors: ["Marcos Casado Barbero", "Romain David", "Iris Van Dam", "Rudolf Wittner"]
+contributors: ["Marcos Casado Barbero", "Romain David", "Iris Van Dam", "Rudolf Wittner", "Enrique Sapena Ventura"]
 page_id: fd_data_description
 redirect_from: /fair-data/data-description
 rdmkit:
@@ -50,10 +50,11 @@ Moreover, by making infectious disease data more findable, researchers promote t
 
 ### Considerations
 
-* Use (globally) unique and persistent [identifiers](https://rdmkit.elixir-europe.org/identifiers) (e.g. [biosample:SAMEA6864906](https://www.ebi.ac.uk/biosamples/samples/SAMEA6864906)) for each of your records, asserting they are unambiguously resolvable from anywhere in the world.
-* Use standard naming conventions for human and disease data (e.g. [Brill-Zinsser](https://www.ebi.ac.uk/ols/ontologies/efo/terms?short_form=EFO_0007182) disease), as well as for taxonomic classifications (e.g. [taxonomy:9606](https://www.ncbi.nlm.nih.gov/Taxonomy/Browser/wwwtax.cgi?mode=Info&id=9606) for humans or [taxonomy:2697049](https://identifiers.org/taxonomy:2697049) for COVID-19). 
-* Describe your data with clear variable names with possible searchable keywords and comprehensive descriptions: choose field standards if possible. Prioritise primary and usual users' standards, but do not forget that metadata may be used by novel users to the field, to which you can cater with generic and understandable. Metadata must be sufficient and appropriate.
-* Register your data as open and accessible as possible through repositories and data portals (e.g. {% tool "ega" %}, {% tool "dbgap" %}, {% tool "health-portal" %}…). 
+* Deidentify patients by means of pseudonymization or anonymization. Examples of software tools for such use are the {% tool "orion-pseudonymization-tool" %} or, for anonymization, {%tool "veil-ai" %}.
+* For samples use (globally) unique and persistent [identifiers](https://rdmkit.elixir-europe.org/identifiers) (e.g. [biosample:SAMEA6864906](https://www.ebi.ac.uk/biosamples/samples/SAMEA6864906)) for each of your records, asserting they are unambiguously resolvable from anywhere in the world.
+* Use standard naming conventions for human and disease data (e.g. [SNOMED](https://www.snomed.org) or [ICD11](https://icd.who.int/en/)), as well as for taxonomic classifications (e.g. [taxonomy:9606](https://www.ncbi.nlm.nih.gov/Taxonomy/Browser/wwwtax.cgi?mode=Info&id=9606) for humans or [taxonomy:2697049](https://identifiers.org/taxonomy:2697049) for COVID-19). 
+* Describe your data with clear variable names with possible searchable keywords and comprehensive descriptions: choose field standards if possible. Several metadata standards have been developed for different diseases. Metadata ontologies can be found for example at {% tool "bioportal" %} or {% tool "ols" %}. Prioritise primary and usual users' standards, but do not forget that metadata may be used by novel users to the field, to which you can cater with generic and understandable. Metadata must be sufficient and appropriate.
+* Submit and register your data as open and accessible as possible through repositories and data portals (e.g. {% tool "ega" %}, {% tool "dbgap" %}, {% tool "health-portal" %}…). 
 * Make your webpage machine accessible and readable, especially for search engines. You can always check the findability of the data you submitted (e.g. using a new session on a web browser), adjust and correct it if needed.
 
 ### Existing approaches
@@ -64,16 +65,16 @@ Moreover, by making infectious disease data more findable, researchers promote t
 
 ## Accessibility
 
-Accessibility in infectious diseases research is crucial to ensure quick, secure and equitable access to clinical and biomolecular data for all, regardless of the reputational or economical power of researchers and institutions, thus promoting the infectious disease research to be community driven.
+Data accessibility in infectious diseases research is crucial to ensure quick, secure and equitable access to clinical and biomolecular data for all, regardless of the reputational or economical power of researchers and institutions, thus promoting the infectious disease research to be community driven.
 
 ### Considerations
 
-* Albeit the emergency of an infectious disease outbreak may pose a tempting chance to obtain human (meta)data through a back-door, when dealing with personal data, and specially [sensitive information](https://rdmkit.elixir-europe.org/sensitive_data), privacy and security measures must remain in place:
+* Despite the emergency of action during an outbreak of an infectious disease at any scale, security measures for handling huma personal data, in particular sensitive data, must remain in place. General Data Protection Regulations (GDPR) must be considered at all times.
     * Sensitive (meta)data must be accessible through the authorised (and community approved) procedures of the archives to which they are submitted. 
-    * You should avoid distributing (meta)data through unsecured channels (e.g. sending metadata of patients in an email to another researcher elsewhere), and instead make use of the procedures already in place by archives. 
+    * Do not distribute (meta)data through unsecured channels (e.g. sending metadata of patients in an email to another researcher elsewhere). If in doubt, consult with IT experts  first to discuss and implement the safest way of sharing data.
 * In a rapid environment like an infectious disease outbreak, (meta)data may also change rapidly: human errors may occur and making changes to (meta)data records may be a necessity. Nevertheless, for the sake of traceability, it is crucial that the (meta)data you submit to an archive is not removed, even when the information it contains is technically wrong: there are other alternatives (e.g. record deprecation) that allow for traceability. Additionally, it is also important to state when a record has been withdrawn, updated or replaced.
 * For all of these aspects, your institution will often do the work for you, you just need to follow their guidance. If not existing, report your action to national repositories or international disciplinary repositories: they provide efficient support for data deposit.
-* Check the accessibility of your data. Your dataset, when submitted to an archive, is likely to have a landing page with several data access services (e.g., download, transcription, contact person, visualisation…). From the perspective of someone foreign to the system to access the data, try to find out how easy it is and then adjust and retest. To increase the quality of accessibility, you can also apply the stable [W3C WCAG](https://www.w3.org/WAI/standards-guidelines/wcag/) and use their tools and methods to assess it.
+* Check the accessibility of your data. Your dataset, when submitted to an archive, is likely to have a landing page with several data access services (e.g., download, transcription, contact person, visualisation…). From the perspective of someone foreign to the system to access the data, try to find out how easy it is and then adjust and retest. To increase the quality of accessibility, you can apply the stable [W3C WCAG](https://www.w3.org/WAI/standards-guidelines/wcag/) guidelines and use their tools and methods to assess it, or contact the host organization and collaborate with them on any issue.
 
 ### Existing approaches 
 
@@ -95,7 +96,7 @@ Without interoperability, data silos may emerge, restricting researchers' abilit
 * Make use of existing metadata standards, such as Data Catalog Vocabularies (DCATs) or the Minimum information about a single amplified genome (MISAG) and a metagenome-assembled genome (MIMAG) of bacteria and archaea ([Bowers et al., 2017](https://www.nature.com/articles/nbt.3893)).
 * Structure your data so that it is [machine-actionable](https://rdmkit.elixir-europe.org/machine_actionability#what-does-machine-readable-machine-actionable-or-machine-interpretable-mean-for-data-and-metadata-in-rdm).
 * Your data should include qualified references to other data sources and metadata, which would increase the traceability and context of your dataset. This may ultimately be needed for necessary meta analyses in pandemic situations. References in fields like the source of patient data (e.g. [UBERON:0001707](http://purl.obolibrary.org/obo/UBERON_0001707)), the laboratory that performed the analysis (e.g. including the name of the laboratory, the name of the institution, and its location), or the specific protocol (e.g. [sample collection](https://journals.asm.org/doi/full/10.1128/JCM.40.11.3956-3963.2002)) used, greatly enhance the quality and transparency of the data. 
-* Verify interoperability by reviewing the definitions of variables and linking them to relevant ontology concepts, ensuring that potential users can assess the compatibility of their data. To achieve this, import corresponding definitions and consult with your data producers to confirm that field definitions do not have significant discrepancies with the data you generate. That has to be done each time you integrate new variables and variable names/definitions.
+* Verify interoperability by reviewing the definitions of variables and linking them to relevant ontology concepts, ensuring that potential users can assess the compatibility of their data. To achieve this, import corresponding definitions and consult with your data producers to confirm that field definitions do not have any discrepancies with the data you generate. That has to be done each time you integrate new variables and variable names/definitions.
 
 ### Existing approaches
 
@@ -112,9 +113,10 @@ Infectious disease research heavily relies on the reusability of human clinical
 
 ### Considerations
 
-* Metadata quality is needed to truly see the data quality and data reusability. Without the metadata, the data, regardless of its quality, is unusable.
+* Metadata quality and completeness is needed to truly see the data quality and data reusability. Without the metadata, the data, regardless of its quality, is unusable. For more information please refer to [RDMkit's metadata documentation](https://rdmkit.elixir-europe.org/metadata_management), [ELIXIR's FAIR CookBook](https://faircookbook.elixir-europe.org/content/recipes/interoperability/creating-minimal-metadata-profiles.html) and [HealthData@EU's metadata validator](https://health-data-itb-rdf-validator.acceptance.data.health.europa.eu/shacl/ehds/upload).
 * The provenance of your data must be transparent for effective tracking of infectious diseases. Key provenance fields include, for example: the geographic location of the study or data source, the date and time of data collection, sources of patient demographic information (e.g., age, gender, ethnicity), sample collection and processing methods (e.g., blood draw, tissue biopsy, RNA extraction), and ethical or legal considerations (e.g., informed consent, data ownership, privacy regulations). The content of provenance information is always context-dependent, tailored to the purpose of provenance collection and the nature of the data produced. Efficient and accurate metadata for provenance should also capture all sensor parameters, ensuring replicability even if the original sensor is no longer available.
-* Make sure that the clinical data you intend to share complies with the applicable laws regarding privacy (e.g. [GDPR Art. 9](https://gdpr-info.eu/art-9-gdpr/)), and rely on the existing data archives for its distribution. When possible, make sure that your data is not completely blocked from being reused, which would render it unusable when shared. Remember: as safe as it needs, but as open for research as the regulations allow.
+* Make sure that the clinical data you intend to share complies with the applicable laws regarding privacy, e.g. [GDPR Art. 9](https://gdpr-info.eu/art-9-gdpr/): In March 2025, the European Health Data Space (EHDS) Regulation entered in force, enforcing common standards and practices aimed at facilitating the reuse of human health data for research. This includes implementing metadata standards that support semantic and technical interoperability. The Health DCAT-AP metadata model, an extension of the DCAT-AP specifically designed for health data, should be used to describe datasets, ensuring consistency in resource description and facilitating data discovery and reuse through national and European data catalogues. By aligning metadata with Health DCAT-AP data providers can support lawful reuse under the EHDS regulation and in accordance with GDPR Art. 9(2)(j).
+* Rely on the existing data archives for its distribution. When possible, make sure that your data is not completely blocked from being reused, which would render it unusable when shared. Remember: as safe as it needs, but as open for research as the regulations allow.
 * Infectious diseases may require an even quicker approach to data discoverability (multi-indexation), distribution and reuse. Within this timeframe, developing a plan from scratch for the data collection, storage, sharing, access, dissemination and reuse seems unfeasible. This timeframe advocates for a thorough preparation: you should create protocols for each of these steps, so that when the time comes, you and your team are prepared.
 * Check your reusability (your dataset should have a landing page with several data and other digital object access services e.g. download, transcription, contact person, visualisation…) -> adjust and retest with different kinds of users to better understand their needs. If potential users are not sure of your data content : data quality, they never reuse them for their own study. (several checklists exist, see for instance: [PARSEC DDOMP](https://doi.org/10.5281/zenodo.3891427))