standards-and-repository-policy.json

{
  "schemaVersion": "1.0",
  "assayPolicies": [
    {
      "assay": "AssayClinicalPrediction",
      "standard": "TRIPOD+AI / PROBAST+AI",
      "scope": "reporting and bias review",
      "depositionTargets": ["repository target review"]
    },
    {
      "assay": "AssayGWAS",
      "standard": "STREGA / PRS-RS",
      "scope": "reporting and ancestry or replication checks",
      "depositionTargets": ["GWAS Catalog", "dbGaP", "EGA"]
    },
    {
      "assay": "AssayBulkRNASeq",
      "standard": "MINSEQE",
      "scope": "RNA-seq reporting",
      "depositionTargets": ["GEO", "SRA", "CNGB CNSA", "NGDC GSA", "ENA"]
    },
    {
      "assay": "AssaySingleCellRNASeq",
      "standard": "HCA Tier-1 metadata schema + scRNA-seq reporting matrix",
      "scope": "donor-level inference checks",
      "depositionTargets": ["GEO", "SRA", "CNGB CNSA", "NGDC GSA", "ENA"]
    },
    {
      "assay": "AssaySpatialTranscriptomics",
      "standard": "spatial transcriptomics reporting",
      "scope": "image provenance checks",
      "depositionTargets": ["GEO", "SRA", "CNGB CNSA", "NGDC GSA", "ENA", "spatial companion repository"]
    },
    {
      "assay": "AssayMicrobiome",
      "standard": "MIxS / STREAMS",
      "scope": "reporting and contamination controls",
      "depositionTargets": ["repository target review"]
    },
    {
      "assay": "AssayImagingOrPathology",
      "standard": "CLAIM / STARD-AI",
      "scope": "diagnostic accuracy reporting",
      "depositionTargets": ["repository target review"]
    },
    {
      "assay": "AssayProteomics",
      "standard": "MIAPE",
      "scope": "proteomics reporting",
      "depositionTargets": ["PRIDE"]
    },
    {
      "assay": "AssayMetabolomics",
      "standard": "MetaboLights reporting expectations",
      "scope": "metabolomics reporting",
      "depositionTargets": ["MetaboLights"]
    },
    {
      "assay": "AssayWGSOrWES",
      "standard": "project-specific reporting matrix",
      "scope": "as determined by project profile",
      "depositionTargets": ["dbGaP", "EGA", "SRA"]
    }
  ],
  "defaultPolicy": {
    "assay": "default",
    "standard": "project-specific reporting matrix",
    "scope": "as determined by project profile",
    "depositionTargets": ["repository target review"]
  },
  "outstandingRequirements": [
    {
      "id": "repository_accession_prerequisites",
      "text": "Confirm repository accession prerequisites and metadata package completeness.",
      "whenPublicationScopeExternal": true,
      "whenBoundedEvidenceSeeking": false,
      "whenOpenEndedBrowsingDefault": false
    }
  ],
  "acceptedReadRepositories": {
    "description": "Repositories the pipeline is permitted to read FROM for raw data acquisition (distinct from depositionTargets, which governs publishing). Agents must NOT drop SME-named accessions solely because a tool only speaks GEO — route by accession prefix to the correct reader.",
    "repositories": [
      {
        "id": "GEO",
        "name": "NCBI Gene Expression Omnibus",
        "accession_prefixes": ["GSE", "GSM", "GDS", "GPL"],
        "preferred_readers": ["GEOparse (Python)", "GEOquery (R/Bioconductor)"],
        "base_url": "https://www.ncbi.nlm.nih.gov/geo/"
      },
      {
        "id": "SRA",
        "name": "NCBI Sequence Read Archive",
        "accession_prefixes": ["SRP", "SRX", "SRR", "PRJNA", "SAMN"],
        "preferred_readers": ["pysradb", "sra-toolkit prefetch"],
        "base_url": "https://www.ncbi.nlm.nih.gov/sra"
      },
      {
        "id": "ENA",
        "name": "European Nucleotide Archive",
        "accession_prefixes": ["ERP", "ERR", "ERX", "PRJEB"],
        "preferred_readers": ["enaBrowserTools", "requests (ENA REST)"],
        "base_url": "https://www.ebi.ac.uk/ena/"
      },
      {
        "id": "CNGB_CNSA",
        "name": "China National GeneBank — CNSA (CNGB Sequence Archive)",
        "accession_prefixes": ["CNP", "CNS", "CNX", "CNR"],
        "preferred_readers": ["HTTPS fetch from ftp.cngb.org/pub/CNSA"],
        "base_url": "https://db.cngb.org/",
        "notes": "CNGB CNSA hosts a substantial fraction of Chinese-authored scRNA-seq IVD and cancer atlases. Always included in the candidate pool for human-tissue scRNA-seq projects — dropping a CNGB-only study reduces cohort coverage without governance justification."
      },
      {
        "id": "NGDC_GSA",
        "name": "National Genomics Data Center (China) — Genome Sequence Archive",
        "accession_prefixes": ["CRA", "HRA", "PRJCA", "GSA"],
        "preferred_readers": ["requests (NGDC download API)", "ngdc_fetch helper"],
        "base_url": "https://ngdc.cncb.ac.cn/gsa/",
        "notes": "NGDC GSA public-access partition. Use HRA (Human Restricted Access) only with DUA; the CRA + open GSA partitions are unrestricted."
      },
      {
        "id": "EGA",
        "name": "European Genome-phenome Archive",
        "accession_prefixes": ["EGAS", "EGAD"],
        "preferred_readers": ["pyega3"],
        "base_url": "https://ega-archive.org/",
        "notes": "EGA is controlled-access by default — only read from EGA when explicit DUA is in the package manifest. Out of scope for public-data-only runs."
      }
    ],
    "acquisitionRoutingRule": "When the SME-named cohort includes accessions from multiple repositories, the data_acquisition candidate method MUST dispatch per-accession-prefix and MUST NOT silently truncate the cohort. Dropping a study because the selected tool can't reach its repository is never a terminal action — surface a structured decision point with options {wait_for_repo_client | defer_to_later_pass | pull_via_alternative_repo_helper}."
  }
}