Semantic metadata and documentation for the National Childhood Cancer Registry Data Platform
The National Childhood Cancer Registry (NCCR) Data Platform is a secure, interactive system for exploring, discovering, and requesting de-identified cancer data for children, adolescents, and young adults (ages 0–39).
This site provides machine-readable metadata for the platform’s datasources and data dictionaries, published as linked data using standard semantic web vocabularies.
| Section | Description |
|---|---|
| Metadata | DATMM catalog records, RDF vocabulary, and instance data |
| Data Sources | Descriptions of the 9 NCCR datasources |
| Vocabulary | Classes and properties defined in the NCCR vocabulary |
| Usage & Examples | Code snippets for Python, R, and SPARQL |
| Cohort Builder Tool | CLI tool for discovering variables, exploring value frequencies, and building cohort definitions |
| Data Cut ERD | Interactive entity-relationship diagram of a full data cut (full-screen) |
| Metadata Graph | Interactive graph of the ontology and example data, parsed live from the Turtle |
| MCP Server | Query NCCR metadata from any AI assistant (Claude, Cursor, VS Code, Kiro) |
| Data Request Skill | Helps a researcher draft and validate an NCCR data request |
Two pieces work together so you can explore NCCR and prepare a data request conversationally:
1. The MCP server — register it once in your AI editor, then ask questions in plain language. It answers from the published metadata, with real record counts and no access to patient data. Setup instructions →
2. The data request skill — drop it into your assistant’s skills folder
(~/.kiro/skills/, ~/.claude/skills/, or your framework’s equivalent) and it guides
you through the NCCR Data Request form: drafting the narrative sections within the
1500-character limits, choosing Research Areas, the SEER Research Plus collaborator
rules, and selecting data elements.
Install instructions →
The server supplies the facts; the skill supplies the judgment. The assistant drafts and validates — you review, attach your IRB and approval documents, submit, and sign.
| File | Description |
|---|---|
| nccr_vocab.ttl | Vocabulary — OWL ontology (schema) |
| nccr_instances.ttl | Instance data — 44,293 triples with value frequencies and patient counts |
| datmm/ | DATMM catalog — 9 standalone dataset records + repository |
Childhood Cancer Data Initiative (CCDI) National Childhood Cancer Registry (NCCR) Data Platform: An interactive data platform for NCCR cancer statistics [Internet]. National Cancer Institute. DOI: 10.71925/byd9-3d93.