Skip to content

competency_questions.md

Latest commit

 

History

History
353 lines (300 loc) · 11.6 KB

competency_questions.md

File metadata and controls

353 lines (300 loc) · 11.6 KB

OKN-WOBD Competency Questions

This document contains competency questions (CQs) for the OKN-WOBD knowledge graph, which represents datasets from the NIAID Data Ecosystem (NDE) including ImmPort, VDJServer, Vivli, RADx Data Hub, and Project Tycho.

Each competency question is written in natural language and includes a corresponding SPARQL query that answers the question.

Dataset Discovery Questions

CQ1: How many datasets are available from ImmPort?

SPARQL Query:

PREFIX schema: <http://schema.org/>
PREFIX rdf: <http://www.w3.org/1999/02/22-rdf-syntax-ns#>

SELECT (COUNT(DISTINCT ?dataset) AS ?count)
WHERE {
    ?dataset rdf:type schema:Dataset .
    FILTER(REGEX(STR(?dataset), "^https://okn\\.wobd\\.org/dataset/immport/"))
}

CQ2: What datasets are available for influenza research?

SPARQL Query:

PREFIX schema: <http://schema.org/>
PREFIX rdf: <http://www.w3.org/1999/02/22-rdf-syntax-ns#>

SELECT DISTINCT ?dataset ?datasetName ?diseaseName
WHERE {
    ?dataset rdf:type schema:Dataset ;
             schema:name ?datasetName ;
             schema:healthCondition ?disease .
    ?disease schema:name ?diseaseName .
    FILTER(
        ?disease = <http://purl.obolibrary.org/obo/MONDO_0005812> ||
        CONTAINS(LCASE(?diseaseName), "influenza")
    )
}
ORDER BY ?datasetName

Note: This query matches datasets by either the MONDO ontology CURIE for influenza (MONDO:0005812) or by string matching on the disease name. If you need a different influenza CURIE, replace MONDO:0005812 with the appropriate identifier.

CQ3: Find all datasets that use mouse as a model organism

SPARQL Query:

PREFIX schema: <http://schema.org/>
PREFIX rdf: <http://www.w3.org/1999/02/22-rdf-syntax-ns#>

SELECT DISTINCT ?dataset ?datasetName ?speciesName
WHERE {
    ?dataset rdf:type schema:Dataset ;
             schema:name ?datasetName ;
             schema:species ?species .
    ?species schema:name ?speciesName .
    FILTER(
        ?species = <https://www.uniprot.org/taxonomy/10090> ||
        REGEX(LCASE(?speciesName), "mouse|mus musculus")
    )
}
ORDER BY ?datasetName

Note: This query matches datasets by either the UniProt taxonomy CURIE for mouse (10090) or by string matching on the species name.

CQ4: What datasets are related to COVID-19?

SPARQL Query:

PREFIX schema: <http://schema.org/>
PREFIX rdf: <http://www.w3.org/1999/02/22-rdf-syntax-ns#>

SELECT DISTINCT ?dataset ?datasetName ?resource
WHERE {
    ?dataset rdf:type schema:Dataset ;
             schema:name ?datasetName ;
             schema:healthCondition <http://purl.obolibrary.org/obo/MONDO_0100096> .
    BIND(REPLACE(STR(?dataset), "https://okn.wobd.org/dataset/([^/]+)/.*", "$1") AS ?resource)
}
ORDER BY ?resource ?datasetName

CQ5: List all datasets with their DOIs

SPARQL Query:

PREFIX schema: <http://schema.org/>
PREFIX rdf: <http://www.w3.org/1999/02/22-rdf-syntax-ns#>

SELECT DISTINCT ?dataset ?datasetName ?doi
WHERE {
    ?dataset rdf:type schema:Dataset ;
             schema:name ?datasetName ;
             schema:sameAs ?doi .
    FILTER(REGEX(STR(?doi), "^https://doi\\.org/"))
}
ORDER BY ?datasetName
LIMIT 50

Funding and Organization Questions

CQ6: Which datasets are funded by NIAID?

SPARQL Query:

PREFIX schema: <http://schema.org/>
PREFIX rdf: <http://www.w3.org/1999/02/22-rdf-syntax-ns#>

SELECT DISTINCT ?dataset ?datasetName ?grantName
WHERE {
    ?dataset rdf:type schema:Dataset ;
             schema:name ?datasetName ;
             schema:funding ?grant .
    ?grant schema:funder <https://ror.org/043z4tv69> ;
           schema:name ?grantName .
}
ORDER BY ?datasetName

CQ7: How many unique funding organizations are represented across all datasets?

SPARQL Query:

PREFIX schema: <http://schema.org/>
PREFIX rdf: <http://www.w3.org/1999/02/22-rdf-syntax-ns#>

SELECT (COUNT(DISTINCT ?funder) AS ?uniqueFunders)
WHERE {
    ?dataset rdf:type schema:Dataset ;
             schema:funding ?grant .
    ?grant schema:funder ?funder .
}

CQ8: Find all datasets authored by researchers from Stanford University

SPARQL Query:

PREFIX schema: <http://schema.org/>
PREFIX rdf: <http://www.w3.org/1999/02/22-rdf-syntax-ns#>

SELECT DISTINCT ?dataset ?datasetName ?authorName ?affiliation
WHERE {
    ?dataset rdf:type schema:Dataset ;
             schema:name ?datasetName ;
             schema:author ?author .
    ?author schema:name ?authorName ;
            schema:affiliation ?org .
    ?org schema:name ?affiliation .
    FILTER(
        # Match by ROR CURIE for Stanford OR by name
        ?org = <https://ror.org/00f54p054> ||
        CONTAINS(LCASE(?affiliation), "stanford")
    )
}
ORDER BY ?datasetName

Note: This query matches datasets by either the ROR identifier for Stanford University (https://ror.org/00f54p054) or by string matching on the affiliation name.

Disease and Infectious Agent Questions

CQ9: What infectious agents are studied across all datasets?

SPARQL Query:

PREFIX schema: <http://schema.org/>
PREFIX rdf: <http://www.w3.org/1999/02/22-rdf-syntax-ns#>

SELECT DISTINCT ?agent ?agentName (COUNT(?dataset) AS ?datasetCount)
WHERE {
    ?dataset rdf:type schema:Dataset ;
             schema:infectiousAgent ?agent .
    ?agent schema:name ?agentName .
}
GROUP BY ?agent ?agentName
ORDER BY DESC(?datasetCount)
LIMIT 20

CQ10: Which datasets study both influenza and use human subjects?

SPARQL Query:

PREFIX schema: <http://schema.org/>
PREFIX rdf: <http://www.w3.org/1999/02/22-rdf-syntax-ns#>

SELECT DISTINCT ?dataset ?datasetName
WHERE {
    ?dataset rdf:type schema:Dataset ;
             schema:name ?datasetName ;
             schema:healthCondition ?disease ;
             schema:species ?species .
    ?disease schema:name ?diseaseName .
    ?species schema:name ?speciesName .
    FILTER(
        # Match influenza by CURIE OR name
        ?disease = <http://purl.obolibrary.org/obo/MONDO_0005812> ||
        CONTAINS(LCASE(?diseaseName), "influenza")
    )
    FILTER(
        # Match human by UniProt taxonomy CURIE OR name
        ?species = <https://www.uniprot.org/taxonomy/9606> ||
        REGEX(LCASE(?speciesName), "human|homo sapiens")
    )
}
ORDER BY ?datasetName

Note: This query matches datasets by ontology CURIEs (MONDO for influenza, UniProt taxonomy for human) or by string matching on names.

Cross-Resource Questions

CQ11: Compare the number of datasets across different resources

SPARQL Query:

PREFIX schema: <http://schema.org/>
PREFIX rdf: <http://www.w3.org/1999/02/22-rdf-syntax-ns#>

SELECT ?resource (COUNT(DISTINCT ?dataset) AS ?datasetCount)
WHERE {
    ?dataset rdf:type schema:Dataset .
    BIND(REPLACE(STR(?dataset), "https://okn.wobd.org/dataset/([^/]+)/.*", "$1") AS ?resource)
}
GROUP BY ?resource
ORDER BY DESC(?datasetCount)

CQ12: Find datasets that are accessible for free

SPARQL Query:

PREFIX schema: <http://schema.org/>
PREFIX rdf: <http://www.w3.org/1999/02/22-rdf-syntax-ns#>

SELECT DISTINCT ?dataset ?datasetName ?resource ?url
WHERE {
    ?dataset rdf:type schema:Dataset ;
             schema:name ?datasetName ;
             schema:url ?url ;
             schema:isAccessibleForFree true .
    BIND(REPLACE(STR(?dataset), "https://okn.wobd.org/dataset/([^/]+)/.*", "$1") AS ?resource)
}
ORDER BY ?resource ?datasetName
LIMIT 50

Temporal Questions

CQ13: What datasets were published or modified in 2024?

SPARQL Query:

PREFIX schema: <http://schema.org/>
PREFIX rdf: <http://www.w3.org/1999/02/22-rdf-syntax-ns#>
PREFIX xsd: <http://www.w3.org/2001/XMLSchema#>

SELECT DISTINCT ?dataset ?datasetName ?dateModified ?datePublished
WHERE {
    ?dataset rdf:type schema:Dataset ;
             schema:name ?datasetName .
    OPTIONAL { ?dataset schema:dateModified ?dateModified . }
    OPTIONAL { ?dataset schema:datePublished ?datePublished . }
    FILTER(
        (BOUND(?dateModified) && REGEX(STR(?dateModified), "^2024")) ||
        (BOUND(?datePublished) && REGEX(STR(?datePublished), "^2024"))
    )
}
ORDER BY DESC(?dateModified) DESC(?datePublished)
LIMIT 50

Advanced Multi-Hop Questions

CQ14: Find all datasets funded by NIAID that study COVID-19 and use human subjects

SPARQL Query:

PREFIX schema: <http://schema.org/>
PREFIX rdf: <http://www.w3.org/1999/02/22-rdf-syntax-ns#>

SELECT DISTINCT ?dataset ?datasetName ?grantName
WHERE {
    ?dataset rdf:type schema:Dataset ;
             schema:name ?datasetName ;
             schema:healthCondition <http://purl.obolibrary.org/obo/MONDO_0100096> ;
             schema:species ?species ;
             schema:funding ?grant .
    ?species schema:name ?speciesName .
    ?grant schema:funder <https://ror.org/043z4tv69> ;
           schema:name ?grantName .
    FILTER(
        # Match human by UniProt taxonomy CURIE OR name
        ?species = <https://www.uniprot.org/taxonomy/9606> ||
        REGEX(LCASE(?speciesName), "human|homo sapiens")
    )
}
ORDER BY ?datasetName

CQ15: What measurement techniques are used across datasets studying influenza?

SPARQL Query:

PREFIX schema: <http://schema.org/>
PREFIX rdf: <http://www.w3.org/1999/02/22-rdf-syntax-ns#>

SELECT DISTINCT ?technique (COUNT(DISTINCT ?dataset) AS ?datasetCount)
WHERE {
    ?dataset rdf:type schema:Dataset ;
             schema:healthCondition ?disease ;
             schema:measurementTechnique ?tech .
    ?disease schema:name ?diseaseName .
    ?tech schema:name ?technique .
    FILTER(
        # Match influenza by MONDO CURIE OR name
        ?disease = <http://purl.obolibrary.org/obo/MONDO_0005812> ||
        CONTAINS(LCASE(?diseaseName), "influenza")
    )
}
GROUP BY ?technique
ORDER BY DESC(?datasetCount)

Note: This query matches datasets by either the MONDO ontology CURIE for influenza or by string matching on the disease name.

Notes

  • All queries use the OKN-WOBD namespace (https://okn.wobd.org/) for datasets
  • External URIs are used for diseases (MONDO), species (UniProt), and organizations (ROR)
  • Some queries may need adjustment based on the actual data loaded into your SPARQL endpoint

Ontology CURIEs Used in Queries

Many queries use OR conditions to match by both ontology CURIEs and string names for better precision and coverage:

  • Influenza: MONDO:0005812
  • Human: UniProt taxonomy 9606 (Homo sapiens)
  • Mouse: UniProt taxonomy 10090 (Mus musculus)
  • Stanford University: ROR 00f54p054
  • COVID-19: MONDO:0100096

Note: If the CURIEs in the queries don't match your data, you can:

  1. Query your data to find the actual CURIEs used
  2. Update the queries with the correct CURIEs
  3. Or remove the CURIE condition and rely on string matching only

Query Compatibility

FRINK-Optimized Queries:

  • These queries are optimized for use with FRINK and use SPARQL 1.1 functions:
    • CONTAINS() for substring matching (CQ2, CQ8, CQ10, CQ15)
    • LCASE() for case-insensitive comparisons (CQ2, CQ3, CQ8, CQ10, CQ14, CQ15)
    • REPLACE() for extracting resource names from URIs (CQ4, CQ11, CQ12)
  • All queries have been tested and validated using scripts/test_competency_queries.py

Protege Compatibility:

  • These queries may not work in Protege due to limited SPARQL 1.1 function support:
    • Protege does not support CONTAINS() - use REGEX() with case-insensitive patterns instead
    • Protege does not support LCASE() - use REGEX() with (?i) flag instead
    • Protege may not support REPLACE() - use UNION with explicit resource matching or remove resource extraction
  • If you need Protege-compatible versions, modify the queries accordingly or use the test script to validate alternative syntax