Evaluation for hra-kg

Evaluator: Cortes et al.

Evaluated on: 2025-08-26

This is a manual evaluation intended to identify potential barriers to reuse.

Evaluation Criteria: This evaluation uses the KG-Registry evaluation rubric as described in Cortes et al. (2025) . The rubric assesses knowledge graphs across multiple dimensions including access, provenance, documentation, maintenance, and fitness for purpose.


Access Level and Types

QuestionAnswerComment
Access to data outside of the knowledge graphYUsers can ask biological questions via queries (Basic Usage: https://cns-iu.github.io/hra-kg-supporting-information/#basic-usage, Advanced Usage: https://github.com/cns-iu/hra-kg-supporting-information/blob/main/README.md)
API or online access to the knowledge graphYCan access by SPARQL endpoint, HRA API, and RESTful-API.
Multiple access options availableYDynamic queries through its SPARQL, RESTful-API, and UIs.
Source code availabilityYGithub (https://github.com/hubmapconsortium/hra-kg)
Downloadable knowledge graphYAvailable on Zenodo and HRA CDN

Section Score: 5/5

Provenance of Nodes and Edges

QuestionAnswerComment
Source list providedYTable 2
Source versions informationYThe type, name, and version of each DO is documented (https://lod.humanatlas.io/asct-b/).
Import dependenciesYIn the hra-do-processor.
Node and edge sourcesYBoth nodes and edges have source information
Edges deduplicationNNot explicit. As each digital object produces its own graph, the triples from different objects may have similar relationships, but they are stored in separate contexts.
Triples source detailsYThe triples are generated by Digital Objects in in RDF format.
Edge type schemaYThe KG is built based on RDF graph (CCF ontology) and standard ontologies to define relationship types between the entities.

Section Score: 6/7

Documented standards, schema, construction

QuestionAnswerComment
Biological usable dataYThe data is human readable and usable in various biological applications.
Resolvable IDsYIt uses resolvable IDs for the entity resolution such as UBERON:0013702 and ENSG00000175899.
Construction documentationYIt provides comprehensive documentation on its construction process (HRA KG Construction and Deployment).
Transformation documentationYIt documents data transfroms such as data normalization and enrichment.
Schema usedYASCT+B tables and Common Coordinate Framework (CCF) Ontology.

Section Score: 5/5

Update frequency and versioning

QuestionAnswerComment
Stable versionsYE.g., v2.3.
Public tracker informationNNot explicitly mentioned, but there are multiple Github Issues portals for users to post requests, reports bugs, and discussions.
Knowledge graph contact informationYContact information is available to reach out to the team (https://humanatlas.io/about#editorial-board)
Updated annuallyYAbout 2 time a year (https://github.com/hubmapconsortium/hra-kg/releases).
Prior versions accessYPrior versions are accessible on Github (https://github.com/hubmapconsortium/hra-kg/releases). Changed are documented (https://humanatlas.io/release-notes/v2.3).

Section Score: 4/5

Evaluation - Metrics and Fitness for Purpose

QuestionAnswerComment
Use case providedYThey have several use cases and applications (Usage Notes).
Evaluation against other modelsYThe HRA-KG is compared with other KGs only from technical perspective, such as the number of nodes, edges, and technologies used.
Defined scopeYHRA-KG is focused on representing the healthy human body in semantics terms including anatomical structures, cell types, and biomarker and spatial terms such as 3D reference organs.
Multiple evaluation methodsYIt has hra-ols used by OLS for validation (https://lod.humanatlas.io/collection/hra-ols/latest/) and has Validation report (https://github.com/hubmapconsortium/3d-hra-ref-object-validation) (Supplementary).
Accuracy metricsNNot explicit, but it has HRAlit which links the HRA DOs to publications, experts, experimental datasets (Supplementary).

Section Score: 4/5

License Information

QuestionAnswerComment
LicenseCC BY 4.0