ICEES Reference Ingest Guide¶
Source Information¶
InfoRes ID: infores:icees-kg
Description: The Integrated Clinical and Environmental Exposures Service (ICEES) is developed primarily by the Renaissance Computing Institute at the University of North Carolina at Chapel Hill as part of the NCATS Biomedical Data Translator. Its goal is to enable secure exploration of relationships between clinical characteristics and environmental exposures while protecting patient privacy, supporting cohort discovery, hypothesis generation, and translational biomedical research. ICEES provides access regulatory-compliant open access integrated, de-identified clinical (e.g., electronic health records, survey data, clinical research data) and environmental exposure (e.g., airborne pollutants, landfills, concentrated animal feeding operations, socio-economic indicators) data through aggregated feature tables and statistical association services rather than individual patient records. The knowledge is generated by integrating electronic health record data with environmental exposure datasets using standardized data models and privacy-preserving aggregation methods. Automated computational pipelines harmonize clinical variables, geospatial exposure measurements, and temporal information, enabling statistical analyses while ensuring that protected health information is not exposed.
Data Access Locations: - https://stars.renci.org/var/plater/bl-4.2.1/icees-kg/8-20-2024/ - Automat ICEES Plater loaded KGX jsonl files, plus meta_knowledge_graph.json
Data Provision Mechanisms: file_download
Data Formats: kgx
Data Versioning and Releases: Static dataset released 8-20-2024
Ingest Information¶
Ingest Categories: aggregation_interpreter
Utility: The impact of environmental exposures on clinical status is well known. This knowledge resource aggregates such data from significant institutional expertise within Translator.
Scope: Clinical correlations to environmental exposures.
Relevant Files¶
| File Name | Location | Description |
|---|---|---|
| nodes.jsonl.gz | https://stars.renci.org/var/plater/bl-4.2.1/icees-kg/8-20-2024 | KGX jsonl file of KGX 'nodes' data from Phase 2 ICEES clinical knowledge provider activities. |
| edges.jsonl.gz | https://stars.renci.org/var/plater/bl-4.2.1/icees-kg/8-20-2024 | KGX jsonl file of KGX 'edges' data from Phase 2 ICEES clinical knowledge provider activities. |
Included Content¶
| File Name | Included Records | Fields Used |
|---|---|---|
| nodes.jsonl | all | id, name, category, equivalent_identifiers |
| edges.jsonl | all | id, subject, predicate, object, primary_knowledge_source, attributes |
Additional Notes: None
Target Information¶
Edge Types¶
| Subject Categories | Predicate | Object Categories | Knowledge Level | Agent Type | UI Explanation |
|---|---|---|---|---|---|
| biolink:Disease, biolink:EnvironmentalExposure, biolink:ClinicalIntervention, biolink:PopulationOfIndividualOrganisms, biolink:PhysiologicalProcess, biolink:ComplexMolecularMixture, biolink:SequenceVariant, biolink:SmallMolecule, biolink:Protein, biolink:MolecularMixture, biolink:PhenotypicFeature, biolink:Drug, biolink:GrossAnatomicalStructure" | biolink:Disease, biolink:EnvironmentalExposure, biolink:ClinicalIntervention, biolink:PopulationOfIndividualOrganisms, biolink:PhysiologicalProcess, biolink:ComplexMolecularMixture, biolink:SequenceVariant, biolink:SmallMolecule, biolink:Protein, biolink:MolecularMixture, biolink:PhenotypicFeature, biolink:Drug, biolink:GrossAnatomicalStructure" | knowledge_assertion | data_analysis_pipeline | Observed relationships between clinical characteristics and environmental exposures, indexed against associated studies. |
Node Types¶
| Node Category | Source Identifier Types | Additional Notes |
|---|---|---|
| biolink:SmallMolecule | UMLS, CHEBI, PUBCHEM.COMPOUND, UNII | |
| biolink:ChemicalExposure | UMLS, CHEBI, PUBCHEM.COMPOUND, UNII | |
| biolink:Drug | UMLS, CHEBI, PUBCHEM.COMPOUND, UNII | |
| biolink:Disease | MONDO, UMLS, NCIT, EFO, UBERON, HP | |
| biolink:InformationContentEntity | UMLS | |
| biolink:PhysiologicalProcess | UMLS | |
| biolink:ComplexMolecularMixture | UMLS | |
| biolink:EnvironmentalExposure | UMLS, PUBCHEM.COMPOUND, NCIT | |
| biolink:PhenotypicFeature | MONDO, UMLS, EFO, NCIT, CHEBI, UBERON, NCBIGene, MESH, NCBITaxon, HP | |
| biolink:Activity | UMLS | |
| biolink:PopulationOfIndividualOrganisms | UMLS | |
| biolink:PhysicalEntity | UMLS | |
| biolink:GrossAnatomicalStructure | UBERON | |
| biolink:OrganismAttribute | UMLS | |
| biolink:ChemicalEntity | MESH, UMLS, CHEBI, NCIT | |
| biolink:Device | UMLS | |
| biolink:Phenomenon | UMLS | |
| biolink:Procedure | UMLS | |
| biolink:MolecularMixture | PUBCHEM.COMPOUND | |
| biolink:SequenceVariant | MESH, UMLS, NCBIGene | |
| biolink:Protein | UMLS, NCBIGene | |
| biolink:Gene | NCBIGene | |
| biolink:OrganismTaxon | NCBITaxon, UMLS | |
| biolink:ClinicalIntervention | UMLS |
Future Modeling Considerations¶
spoq_pattern: The initial ingest of ICEES KGX data only provides a special mapping for CorrelatedGeneToDiseaseAssociation entries, but emits all other edges as undifferentiated Association edges. There was some attempt in coding to leverage facilities of the Biolink Model Toolkit to provide mappings onto other subclasses of Association but this proved to be semantically misleading; however, there may be a role here for some explicit curation of such mappings based on SPOQ content, using more explicit code inside of the ingest. It may also be the case that the Biolink Model needs to define additional Association child classes to specify such correlations.
Additional Notes: None
Provenance Information¶
Contributors: - Richard Bruskiewich - data modeling, code author - Evan Morris: code support - Sierra Moxon: data modeling, code support - Matthew Brush - data modeling, domain expertise