{
  "snapshot": {
    "title": "Carbon-A assembly strands \u2014 proof of concept",
    "root": 2759,
    "assemblies": 33588,
    "taxa": 16976,
    "release": {
      "assemblies": 48167,
      "taxa": 22617,
      "loci": 566340000,
      "bases": 27180000000000
    },
    "snapshot": "Explorer taxonomy and covered-assembly mapping, 7 Oct 2026",
    "prior_annotation": "NCBI GenBank assembly summary, cached 7 Oct 2026; named annotation provider required",
    "refseq": "Exact paired GCF accession match to the project\u2019s 2,055-assembly RefSeq comparison corpus; not all of RefSeq",
    "classification_counts": {
      "1": 4238,
      "0": 29350
    },
    "refseq_matches": 1323,
    "layout": "NCBI taxonomy; branch lengths are for layout, not evolutionary distance. Each strand is one assembly. Equal strand widths encode assembly counts, not gene counts."
  },
  "sources": {
    "taxonomy_repository": "https://huggingface.co/spaces/HuggingFaceBio/genbank-annotation-explorer",
    "taxonomy_repository_commit": "c6b2b4ac351eaf9680e8141d89bf1f6e76633670",
    "genbank_summary": "https://ftp.ncbi.nlm.nih.gov/genomes/ASSEMBLY_REPORTS/assembly_summary_genbank.txt",
    "genbank_summary_cache_date": "2026-10-07",
    "human_metadata": "HuggingFaceBio/genbank-annotations/cds/vertebrate_mammalian/GCA_000001405.29/metadata.json",
    "human_gff3": "HuggingFaceBio/genbank-annotations/cds/vertebrate_mammalian/GCA_000001405.29/annotations.gff3.gz",
    "human_pipeline_commit": "6ffa159",
    "refseq_comparison": "HF_GENERanno_organized/analysis/training_data_stats/refseq_assembly_taxids.csv",
    "locus_counts": "Per-assembly metadata.json / gene_count in HuggingFaceBio/genbank-annotations/cds, retrieved 2026-10-09",
    "atlas_icons": "Original PhyloPic SVGs and taxid assignments from genbank-annotation-explorer/data/icons.json; per-image credits retained in atlas-icons.json",
    "release_totals": "User-supplied values, 2026-10-08"
  },
  "sha256": {
    "taxonomy.sqlite": "88e8b10a82026dfe21a3022ae788e50d82a4ca521f948a29f4a219e111b185b1",
    "common_names.json": "b46396e6db5fc2326617d9dc20e0f219bbbe79ef3b9329ac2e287671ea3c4812",
    "refseq_assembly_taxids.csv": "336a74474136597863c08188d4c8bc3113057e3409e665a1e937838c7b4f071b",
    "human-metadata.json": "a4b825732a128d0745cfc6b0fa0b8c28c1909bd5a6c1f04eec48ee9584d0b76a",
    "human-annotations.gff3.gz": "c7fa0b8df740641718f65f65a0ef7c9d76c14cf537070d5e6222071c36a68087",
    "tree-data.json": "ee3b3b9373bc9dc9109febef8d8027405b0da0d7fb21a0aea69cac06e42e6dad",
    "example-genes.json": "d9640190ee771eb099df5bb1e059a837552bfbdeb734b1daa929d7a76cec98d7",
    "locus-counts.json": "aa2d237238567ec8e7cb15bb3448d629c4f1698ce2733171dee4b6d8f27b4113",
    "atlas-icons.json": "1f4b52b0e58c8663f2a46ac841915774489c18e15dd9ba9a300c47e723854ddb",
    "release-source-audit.json": "6ea3fbec293107aa46b418e988f25d7437badc192146dd17ca9db557265334a0"
  },
  "important_limits": [
    "This is the 33,588-accession explorer subset, not the complete 48,167-assembly release.",
    "Locus bars use published gene_count values for available assemblies; missing measurements are marked, not imputed.",
    "Existing annotation evidence is at assembly level, not evidence that an individual Carbon-A locus is novel.",
    "RefSeq comparison corpus is a filtered project corpus, not all RefSeq.",
    "Green indicates no listed match under the chosen comparison, not confirmed absence of previous annotation."
  ],
  "locus_coverage": {
    "available_assemblies": 17828,
    "shown_assemblies": 33588,
    "missing_counts": 15760,
    "sum_available_loci": 212905534
  },
  "release_source_audit": {
    "retrieved_at": "2026-10-09T14:01:37.145268+00:00",
    "published_gene_file_directories": 18662,
    "status": "Full 48167-assembly release manifest not located; no synthetic rows or imputed counts."
  },
  "annotation_access": {
    "bucket": "HuggingFaceBio/genbank-annotations",
    "paths": "Embedded from per-assembly metadata source paths",
    "generated_example": "GCA_000001405.29; all 4 files downloaded, annotation/sequence hashes match metadata"
  },
  "presentation": {
    "updated_at": "2026-10-12",
    "layout": "Wide title and tree centred on the full viewport, with a narrow left links column and matching right spacing for assembly details. Full available height below the title. Five equal-width buttons under Links:, followed by the legend. On phones, resources follow the full-width diagram.",
    "animation": "4-second sweep with matching assembly info each frame; Space starts/stops, retaining the current strand on stop; final assembly remains selected; no counters or playback button.",
    "data_changed": false,
    "interaction": "Hover previews and highlights strands and assembly details. Click pins; next click anywhere clears. Icons, bars, double-clicks and search do not zoom. Visible Whole tree resets zoom/pan.",
    "title": "Carbon-A: Discovering Genes Across the Tree of Life",
    "subtitle": "Releasing 566M gene candidates for over 22k species",
    "subtitle_source": "Final user-approved release tweet, 8 October 2026: 566 million gene candidates across 22k+ species; rounded full-release figures, not counts of the displayed subset.",
    "legend": "Green: no GenBank annotation listed; blue: GenBank annotation listed. Updates to the RefSeq comparison corpus when that mode is selected.",
    "icons": "Original Genome Atlas silhouettes: denser placement, branch-count-dependent sizes, and thin dark grey contours. Bounds-based spacing and hit targets; no circle backgrounds. Original SVGs, taxid mappings and credits unchanged.",
    "icon_outline": "Dark grey silhouette contour, 0.4 CSS px on small icons and 0.55 CSS px on larger icons.",
    "space": "HuggingFaceBio/carbon-a-release"
  }
}