Version history

1 version on record. Newest first; the live version sits at the top with a live indicator.

  1. Live 6d45cc4f4ee7
    5/17/2026, 4:35:28 PM
    Content snapshot
    {
      "kind": "infographic",
      "prompt": "Successive Allen taxonomy releases redefine 'transcriptomic cell type' at increasingly fine granularity, making cell-type-specific connectivity rules a moving target.",
      "provider": "other",
      "raw_fields": {
        "papers": [
          {
            "n": 1679,
            "doi": "10.1038/nn.4216",
            "value": "49",
            "method": "single-cell RNA-seq (SMART-seq)",
            "metric": "n_transcriptomic_types_V1",
            "n_analyzed": "1679",
            "ci_or_error": "",
            "text_access": "abstract_only",
            "n_definition": "single cells profiled by SMART-seq",
            "scope_region": "single cortical area (VISp)",
            "study_system": "adult mouse primary visual cortex (V1)",
            "taxonomic_level": "fine type (t-type)",
            "scope_population": "all transcriptomically profiled cortical cells",
            "value_source_sentence": "We identified 49 transcriptomic cell types, including 23 GABAergic, 19 glutamatergic and 7 non-neuronal types.",
            "experimental_conditions": "Cre-line targeted FACS-sorted cells, SMART-seq"
          },
          {
            "n": 23822,
            "doi": "10.1038/s41586-018-0654-5",
            "value": "133",
            "method": "single-cell RNA-seq (SMART-seq)",
            "metric": "n_transcriptomic_types_VISp_ALM",
            "n_analyzed": "23822",
            "ci_or_error": "",
            "text_access": "abstract_only",
            "n_definition": "single cells profiled by SMART-seq",
            "scope_region": "two cortical areas",
            "study_system": "adult mouse primary visual cortex (VISp) and anterior lateral motor cortex (ALM)",
            "taxonomic_level": "fine type (t-type)",
            "scope_population": "transcriptomically profiled cortical neurons",
            "value_source_sentence": "We define 133 transcriptomic cell types by deep, single-cell RNA sequencing.",
            "experimental_conditions": "Cre-line targeted FACS-sorted cells, SMART-seq"
          },
          {
            "n": 1300000,
            "doi": "10.1016/j.cell.2021.04.021",
            "value": "~1,300,000",
            "method": "single-cell RNA-seq (10x)",
            "metric": "n_cells_profiled_isocortex_HPF",
            "n_analyzed": "1300000",
            "ci_or_error": "",
            "text_access": "abstract_only",
            "n_definition": "single cells profiled by scRNA-seq",
            "scope_region": "whole isocortex + hippocampal formation",
            "study_system": "adult mouse isocortex and hippocampal formation",
            "taxonomic_level": "fine type (t-type)",
            "scope_population": "all dissociated cells",
            "value_source_sentence": "We profiled ~1.3 million cells covering the entire adult mouse isocortex and HPF and derived a transcriptomic cell-type taxonomy revealing a comprehensive repertoire of glutamatergic and GABAergic neuron types.",
            "experimental_conditions": "whole isocortex+HPF dissociation, 10x scRNA-seq"
          },
          {
            "n": 4000000,
            "doi": "10.1038/s41586-023-06812-z",
            "value": "5322",
            "method": "scRNA-seq + MERFISH",
            "metric": "n_clusters_whole_brain_atlas",
            "n_analyzed": "4000000",
            "ci_or_error": "",
            "text_access": "abstract_only",
            "n_definition": "single cells profiled by scRNA-seq post-QC",
            "scope_region": "whole mouse brain",
            "study_system": "adult mouse whole brain",
            "taxonomic_level": "cluster",
            "scope_population": "all dissociated cells",
            "value_source_sentence": "The atlas is hierarchically organized into 4 nested levels of classification: 34 classes, 338 subclasses, 1,201 supertypes and 5,322 clusters.",
            "experimental_conditions": "10x scRNA-seq + MERFISH spatial transcriptomics"
          }
        ],
        "audit_issues": [
          {
            "dimension": "scope_region",
            "description": "Rows span V1 only → VISp+ALM → whole isocortex+HPF → whole brain. With expanding brain coverage, cluster counts mechanically increase from added regions, not from finer resolution within a fixed scope. This is the canonical scope-mismatch trap.",
            "entries_affected": [
              "10.1038/nn.4216",
              "10.1038/s41586-018-0654-5",
              "10.1016/j.cell.2021.04.021",
              "10.1038/s41586-023-06812-z"
            ]
          },
          {
            "dimension": "metric_definition",
            "description": "Row 3 reports n_cells_profiled (~1,300,000) — a sample-size measure — while rows 1, 2, 4 report n_transcriptomic_types (49, 133, 5,322). These are different quantities on different scales (cells vs. types) and cannot share a y-axis.",
            "entries_affected": [
              "10.1016/j.cell.2021.04.021"
            ]
          },
          {
            "dimension": "taxonomic_level",
            "description": "Rows mix t-types (rows 1–2) with hierarchically-organised 'clusters' (row 4, finest level of a 4-level taxonomy: 34 classes / 338 subclasses / 1,201 supertypes / 5,322 clusters). 5,322 clusters and 49 t-types are not at the same taxonomic grain.",
            "entries_affected": [
              "10.1038/nn.4216",
              "10.1038/s41586-018-0654-5",
              "10.1038/s41586-023-06812-z"
            ]
          }
        ],
        "audit_verdict": "REDESIGN",
        "comparison_id": "transcriptomic-cell-types-mouse-cortex",
        "comparison_name": "Number of transcriptomic cell types defined in successive mouse cortex single-cell taxonomies",
        "comparison_type": "timeline",
        "what_it_reveals": "Successive Allen taxonomy releases redefine 'transcriptomic cell type' at increasingly fine granularity, making cell-type-specific connectivity rules a moving target.",
        "homogeneity_check": {
          "caveats": [
            "Each study expands geographic coverage (V1 → VISp+ALM → isocortex+HPF → whole brain), so cluster counts are not directly comparable.",
            "Taxonomic granularity differs (t-types vs 'supertypes' vs 'clusters')."
          ],
          "n_definition_uniform": "true",
          "scope_region_uniform": "false",
          "taxonomic_level_uniform": "false",
          "scope_population_uniform": "true"
        },
        "suggested_plot_type": "timeline",
        "mandatory_caption_caveats": [
          "Brain coverage expands from V1 → 2 cortical areas → isocortex+HPF → whole brain across rows; cluster counts are not directly comparable as a 'discovery progression' because they reflect added tissue, not finer resolution.",
          "Taxonomic grain differs: rows 1–2 report t-types; row 4 reports 'clusters' at the finest of four nested levels (whole-brain atlas also has 34 classes / 338 subclasses / 1,201 supertypes that are closer in grain to the earlier t-type counts).",
          "Phase 7 writer: REDESIGN verdict — implement as: Restrict to a single scope (e.g., mouse cortex) and report taxonomic counts at the equivalent level across studies. For the whole-brain atlas (Yao 2023), use the subclass/supertype count rather than the cluster count if comparing to earlier t-type-level papers. Correct the row-3 metric: replace n_ce"
        ]
      },
      "section_id": "section_07",
      "source_url": "https://github.com/AllenNeuralDynamics/ComputationalReviewRecurrence/blob/79ce062d54a924ce05953ec90aa9d26044d2b48f/evidence/section_07_evidence_package.json",
      "target_ref": "wiki_page:computationalreviewrecurrence-07-celltype-motifs",
      "review_repo": "ComputationalReviewRecurrence",
      "section_ref": "wiki_page:computationalreviewrecurrence-07-celltype-motifs",
      "source_path": "evidence/section_07_evidence_package.json",
      "source_refs": [
        "paper:paper-ed2a7f08ef35",
        "paper:paper-pm-30382198",
        "paper:paper-4dfe44516146",
        "paper:1f98e15e-de06-42f7-9da8-1faa69bba835"
      ],
      "section_title": "7. Cell-type-specific E→E motifs in mouse — IT vs PT vs CT pyramidal projection classes; L5 thick-tufted recurrence; Patch-seq and Allen mouse-cortex taxonomy intersections; transcriptomic-type-specific connectivity",
      "source_policy": {
        "mode": "public_source_pointer_with_short_context",
        "notes": [
          "Local review repositories are read-only inputs.",
          "SciDEX stores paper metadata, structured evidence, file pointers, and short citation contexts; it does not copy full review prose."
        ],
        "source_commit_sha": "79ce062d54a924ce05953ec90aa9d26044d2b48f",
        "source_repository_url": "https://github.com/AllenNeuralDynamics/ComputationalReviewRecurrence"
      },
      "generation_status": "complete",
      "review_bundle_ref": "analysis_bundle:ab-d9c479db9be9",
      "origin_url": "https://github.com/AllenNeuralDynamics/ComputationalReviewRecurrence/blob/79ce062d54a924ce05953ec90aa9d26044d2b48f/evidence/section_07_evidence_package.json",
      "commit_sha": "79ce062d54a924ce05953ec90aa9d26044d2b48f",
      "created_by": "persona-jerome-lecoq-gbo-neuroscience",
      "repository_url": "https://github.com/AllenNeuralDynamics/ComputationalReviewRecurrence"
    }