{
  "registry": "SIMARC data commons",
  "version": "v1",
  "updated": "2026-08-21",
  "policy": {
    "cost": "free, no account, no data-use agreement, no fee",
    "bundling": "nothing is bundled; every byte re-downloads from a primary source",
    "rationale": "A dataset you cannot re-fetch is a dataset whose provenance you are asserting rather than demonstrating.",
    "redistributable": "mirrored with SHA-256 manifests",
    "encumbered": "shipped as a re-download script plus checksums - equally reproducible, legally clean",
    "nomination": "request_data(name, source_url, why) over MCP, or contact@simarc.org"
  },
  "datasets": [
    {
      "id": "tcga-pancancer-rnaseq",
      "name": "TCGA pan-cancer RNA-seq",
      "program": "P-03",
      "scale": "33 projects, ~11,500 patients",
      "rung_reachable": "OBSERVATIONAL",
      "access": "fetch + hash",
      "source_url": "https://portal.gdc.cancer.gov/",
      "license": "open access tier",
      "note": "The pilot touched one of 33 projects. The features already downloaded are unused."
    },
    {
      "id": "depmap-crispr",
      "name": "DepMap CRISPR gene effect",
      "program": "P-03",
      "scale": "1,208 cell lines x 18,531 genes",
      "rung_reachable": "PERTURBATIONAL",
      "access": "fetch + hash",
      "source_url": "https://depmap.org/portal/download/",
      "license": "CC BY 4.0",
      "note": "The qualitative jump from correlation to causation. This is the binding constraint, not more DNA."
    },
    {
      "id": "dose-response",
      "name": "Dose-response compendium (CTRPv2, GDSC and others)",
      "program": "P-03",
      "scale": "5,450,000 measurements",
      "rung_reachable": "PHARMACOLOGICAL",
      "access": "fetch + hash",
      "source_url": "https://depmap.org/portal/download/",
      "license": "varies by source, all open"
    },
    {
      "id": "clinicaltrials-oncology",
      "name": "ClinicalTrials.gov - oncology interventional",
      "program": "P-03",
      "scale": "97,623 interventional cancer trials",
      "rung_reachable": "CLINICAL",
      "access": "live query, timestamped",
      "source_url": "https://clinicaltrials.gov/data-api/api",
      "license": "public domain"
    },
    {
      "id": "clinicaltrials-dementia",
      "name": "ClinicalTrials.gov - dementia interventional (the graveyard)",
      "program": "P-04",
      "scale": "8,224 trials, 731 classified dead, 59 evidence-bearing",
      "rung_reachable": "CLINICAL",
      "access": "live query, timestamped",
      "source_url": "https://clinicaltrials.gov/data-api/api",
      "license": "public domain",
      "note": "The obvious query returns 5,242; the field-vocabulary query returns 8,224. Classifier passes 44 controls in both directions."
    },
    {
      "id": "bellenguez-2022-gwas",
      "name": "Bellenguez 2022 Alzheimer's disease GWAS",
      "program": "P-04",
      "scale": "21,101,114 variants, n = 487,511",
      "rung_reachable": "GENETIC",
      "access": "fetch + md5 verified on every run",
      "source_url": "https://www.ebi.ac.uk/gwas/",
      "license": "open, no data-use agreement",
      "checksum_policy": "md5 verified against the EBI published checksum each load"
    },
    {
      "id": "europepmc-dementia",
      "name": "Europe PMC - dementia literature corpus",
      "program": "P-04",
      "scale": "227,231 records on 'Alzheimer disease', with retraction metadata",
      "rung_reachable": "prior art",
      "access": "live query",
      "source_url": "https://europepmc.org/RestfulWebService",
      "license": "open"
    },
    {
      "id": "open-targets-gnomad",
      "name": "Open Targets and gnomAD",
      "program": "P-03, P-04",
      "scale": "associations, drugs, mechanisms, constraint, allele frequencies",
      "rung_reachable": "prior art",
      "access": "fetch + hash",
      "source_url": "https://platform.opentargets.org/",
      "license": "CC0 / open"
    },
    {
      "id": "cmb-skymaps",
      "name": "Planck and WMAP CMB sky maps",
      "program": "P-01",
      "scale": "4 independent sky-cleanings across two missions",
      "rung_reachable": "the boundary seam (ANOMALY)",
      "access": "fetch + hash",
      "source_url": "https://pla.esac.esa.int/",
      "license": "open"
    },
    {
      "id": "quasar-catalog",
      "name": "Quasar catalogs",
      "program": "P-01",
      "scale": "756,000 objects",
      "rung_reachable": "the lattice seam (NULL)",
      "access": "fetch + hash",
      "source_url": "https://www.sdss.org/dr17/",
      "license": "open"
    },
    {
      "id": "butson-corpus-6",
      "name": "The Butson corpus at order 6",
      "program": "P-01",
      "scale": "27 classes, byte-verified",
      "rung_reachable": "PROOF",
      "access": "mirrored",
      "license": "redistributable",
      "note": "On disk for the MUB(6) question: does the quantum substrate carry a scar at exactly the size where the classical one broke?"
    },
    {
      "id": "drat-proofs",
      "name": "DRAT proof artifacts, order-six program",
      "program": "P-01",
      "scale": "145 MB, hash-pinned, replayable",
      "rung_reachable": "PROOF",
      "access": "mirrored",
      "license": "redistributable",
      "note": "Replayed twice under a hash-pinned checker. A referee can re-check without trusting our code or re-running our solvers."
    },
    {
      "id": "math-subject-surveys",
      "name": "Mathematical subject surveys",
      "program": "P-01",
      "scale": "33 fields",
      "rung_reachable": "reference",
      "access": "mirrored",
      "license": "redistributable"
    }
  ],
  "blocked": [
    {
      "name": "ADNI, NACC, ROSMAP, UK Biobank",
      "program": "P-04",
      "why_blocked": "All require data-use applications. They are where individual-level longitudinal data lives.",
      "consequence": "The lag gate and the competing-risk gate are built, tested in both directions, and currently pointed at no data.",
      "status": "named as blocked rather than worked around; collaboration sought"
    }
  ]
}