cBioPortal governed baseline
The study's clinical schema, governed - headers only, so no patient data ever leaves your site.
The plain on-ramp for the cBioPortal connector (Import, Audit, Generate). The artifacts are the staging files you already curate - only the four # header rows and the attribute-ID row matter (data rows are ignored), so a schema-only upload works and is inherently PHI-safe. Import turns each clinical file into a governed instrument: attributes become Elements with datatypes, descriptions and priorities; PATIENT_ID/SAMPLE_ID become identities; the sample file's PATIENT_ID becomes a governed reference to the patient instrument - the study join, in the graph. On its own this overlaps with what validateData.py tells you per study; its value is what it unlocks: the problem recipes (attribute registry, staging preflight, study-update gate, consortium conformance) all start from this baseline.