{
  "name": "GlioState data dictionary",
  "generated": "2026-10-15",
  "site": "https://gliostate.com",
  "license_note": "Content for research and education only; not medical advice.",
  "counts": {
    "items": 139,
    "topics": 8,
    "entities": 78,
    "entities_by_type": {
      "cell_state": 5,
      "gene": 8,
      "pathway": 3,
      "method": 19,
      "drug": 1,
      "model": 25,
      "dataset": 10,
      "disease": 7
    },
    "relations": 51,
    "relations_by_predicate": {
      "promotes": 2,
      "suppresses": 1,
      "marks": 4,
      "transitions_to": 3,
      "validated_by": 1,
      "benchmarked_on": 5,
      "uses_dataset": 3,
      "perturbs": 2,
      "associated_with": 30
    },
    "tags_total": 60,
    "tags_with_pages": 60,
    "graph_nodes": 291,
    "graph_edges": 1099
  },
  "endpoints": {
    "/items.json": "all items, normalised (array)",
    "/items.csv": "original CSV",
    "/graph.json": "nodes + typed edges",
    "/graph.jsonld": "JSON-LD export of the same graph",
    "/api/items/<VCxxxx>.json": "one item with backlinks and related items",
    "/api/topics/<topic-id>.json": "one research topic (full record + resolved references, entities, typed relations)",
    "/api/entities/<slug>.json": "one entity: description, aliases, external ids, evidence, mentions (items, topics), typed relations (both directions) and related entities",
    "/entities.json": "all entities (array, brief records)",
    "/data/entities/<slug>.json": "raw entity source records",
    "/entities/ · /entities/<slug>/": "entity index and entity HTML pages (zh under /zh/entities/)",
    "/data/graph/<file>.json": "raw research topic source records",
    "/schema.json": "this file",
    "/library/<VCxxxx>/": "paper HTML page",
    "/library/tag/<slug>/ · /library/type/<type>/ · /library/year/<year>/": "facet pages",
    "/library/facets/": "facet index"
  },
  "item": {
    "fields": {
      "id": "string — stable GlioState id (VC0001…)",
      "type": "enum — conference | dataset | interview | model | news | org | paper | preprint",
      "title": "string",
      "authors_or_org": "string — authors (papers) or organisation",
      "date": "string — YYYY-MM-DD, YYYY-MM or YYYY; may be empty",
      "year": "string|null — derived from date",
      "venue": "string",
      "url": "string — source URL (may equal the DOI URL)",
      "doi": "string — bare DOI, no URL prefix",
      "pmid": "string",
      "brain_tumor_focus": "enum — yes | partial | no",
      "topic_tags": "string — ORIGINAL semicolon-separated tags (items.json only; kept for backward compatibility)",
      "tags": "string[] — original tags split into an array (trimmed, composite ;-tags split)",
      "tag_slugs": "string[] — canonical tag slugs from data/tag_vocab.json (aliases merged, e.g. scRNA/scRNA-seq/single-cell → scrna-seq)",
      "tag_labels": "string[] — English labels for tag_slugs",
      "tag_labels_zh": "string[] — Chinese labels for tag_slugs",
      "unmapped_tags": "string[] — raw tags with no controlled-vocabulary entry (no facet page)",
      "one_line_summary_en": "string",
      "one_line_summary_zh": "string",
      "verified": "enum — yes | partial",
      "verification_source": "string — e.g. Europe PMC, PubMed, Crossref, official page",
      "page_url": "string — canonical HTML page /library/<id>/",
      "api_url": "string",
      "sameAs": "string[] — DOI / PubMed / source URLs",
      "is_preprint": "boolean",
      "facets": "object — type, year and tag facet page URLs",
      "cited_by": "object[] — research topics citing this item (backlinks)",
      "related": "object[] — up to 6 related items by shared tags (IDF-weighted)"
    },
    "enums": {
      "type": [
        "conference",
        "dataset",
        "interview",
        "model",
        "news",
        "org",
        "paper",
        "preprint"
      ],
      "brain_tumor_focus": [
        "yes",
        "partial",
        "no"
      ],
      "verified": [
        "yes"
      ]
    }
  },
  "tags": {
    "normalisation": [
      "split on \";\" (and \"/\")",
      "trim and de-duplicate per item",
      "match case-insensitively against the controlled vocabulary data/tag_vocab.json (60 canonical tags with en/zh labels and aliases)",
      "unmapped raw tags are kept as plain text (items.json: unmapped_tags) and get no facet page",
      "facet pages only for canonical tags used by ≥2 items"
    ],
    "vocabulary_size": 60,
    "facet_tags": [
      {
        "slug": "gbm",
        "label": "GBM",
        "label_zh": "胶质母细胞瘤",
        "count": 21
      },
      {
        "slug": "perturbation",
        "label": "Perturbation",
        "label_zh": "扰动",
        "count": 21
      },
      {
        "slug": "spatial",
        "label": "Spatial omics",
        "label_zh": "空间组学",
        "count": 20
      },
      {
        "slug": "foundation-model",
        "label": "Foundation model",
        "label_zh": "基础模型",
        "count": 17
      },
      {
        "slug": "scrna-seq",
        "label": "scRNA-seq",
        "label_zh": "单细胞 RNA 测序",
        "count": 13
      },
      {
        "slug": "benchmark",
        "label": "Benchmark & validation",
        "label_zh": "基准评测与验证",
        "count": 12
      },
      {
        "slug": "conference-meeting",
        "label": "Conferences & workshops",
        "label_zh": "会议与研讨会",
        "count": 10
      },
      {
        "slug": "aivc",
        "label": "AI virtual cell",
        "label_zh": "AI 虚拟细胞",
        "count": 7
      },
      {
        "slug": "drug-response",
        "label": "Drug response",
        "label_zh": "药物反应",
        "count": 7
      },
      {
        "slug": "dataset",
        "label": "Dataset",
        "label_zh": "数据集",
        "count": 6
      },
      {
        "slug": "generative",
        "label": "Generative AI & diffusion",
        "label_zh": "生成式 AI 与扩散模型",
        "count": 6
      },
      {
        "slug": "glioma",
        "label": "Glioma",
        "label_zh": "胶质瘤",
        "count": 6
      },
      {
        "slug": "mechanistic",
        "label": "Mechanistic & agent-based model",
        "label_zh": "机制与基于主体的模型",
        "count": 6
      },
      {
        "slug": "arc",
        "label": "Arc Institute",
        "label_zh": "Arc 研究所",
        "count": 5
      },
      {
        "slug": "challenge",
        "label": "Challenge",
        "label_zh": "挑战赛",
        "count": 5
      },
      {
        "slug": "critique",
        "label": "Critique",
        "label_zh": "批评与反证",
        "count": 5
      },
      {
        "slug": "definition",
        "label": "Definition",
        "label_zh": "定义与概念",
        "count": 5
      },
      {
        "slug": "digital-twin",
        "label": "Digital twin",
        "label_zh": "数字孪生",
        "count": 5
      },
      {
        "slug": "organoid",
        "label": "Organoid",
        "label_zh": "类器官",
        "count": 5
      },
      {
        "slug": "perturb-seq",
        "label": "Perturb-seq",
        "label_zh": "Perturb-seq 扰动测序",
        "count": 5
      },
      {
        "slug": "plasticity",
        "label": "Plasticity",
        "label_zh": "细胞可塑性",
        "count": 5
      },
      {
        "slug": "agents",
        "label": "AI agents",
        "label_zh": "AI 智能体",
        "count": 4
      },
      {
        "slug": "atlas",
        "label": "Atlas",
        "label_zh": "细胞图谱",
        "count": 4
      },
      {
        "slug": "brain-metastasis",
        "label": "Brain metastasis",
        "label_zh": "脑转移瘤",
        "count": 4
      },
      {
        "slug": "cancer",
        "label": "Cancer",
        "label_zh": "癌症",
        "count": 4
      },
      {
        "slug": "cross-species",
        "label": "Cross-species",
        "label_zh": "跨物种",
        "count": 4
      },
      {
        "slug": "czi",
        "label": "CZI",
        "label_zh": "陈-扎克伯格倡议 (CZI)",
        "count": 4
      },
      {
        "slug": "dmg",
        "label": "DMG",
        "label_zh": "弥漫性中线胶质瘤",
        "count": 4
      },
      {
        "slug": "in-silico-perturbation",
        "label": "In-silico perturbation",
        "label_zh": "计算机模拟扰动",
        "count": 4
      },
      {
        "slug": "interview",
        "label": "Interview & podcast",
        "label_zh": "访谈与播客",
        "count": 4
      },
      {
        "slug": "mri",
        "label": "MRI",
        "label_zh": "磁共振成像",
        "count": 4
      },
      {
        "slug": "pathology",
        "label": "Pathology",
        "label_zh": "病理",
        "count": 4
      },
      {
        "slug": "platform",
        "label": "Platform",
        "label_zh": "平台",
        "count": 4
      },
      {
        "slug": "review",
        "label": "Review",
        "label_zh": "综述",
        "count": 4
      },
      {
        "slug": "xaira",
        "label": "Xaira",
        "label_zh": "Xaira",
        "count": 4
      },
      {
        "slug": "genbio",
        "label": "GenBio / AIDO",
        "label_zh": "GenBio / AIDO",
        "count": 3
      },
      {
        "slug": "grn",
        "label": "Gene regulatory network",
        "label_zh": "基因调控网络",
        "count": 3
      },
      {
        "slug": "llm",
        "label": "LLM",
        "label_zh": "大语言模型",
        "count": 3
      },
      {
        "slug": "medulloblastoma",
        "label": "Medulloblastoma",
        "label_zh": "髓母细胞瘤",
        "count": 3
      },
      {
        "slug": "meningioma",
        "label": "Meningioma",
        "label_zh": "脑膜瘤",
        "count": 3
      },
      {
        "slug": "multiscale",
        "label": "Multiscale",
        "label_zh": "多尺度",
        "count": 3
      },
      {
        "slug": "news",
        "label": "News",
        "label_zh": "新闻",
        "count": 3
      },
      {
        "slug": "noetik",
        "label": "Noetik",
        "label_zh": "Noetik",
        "count": 3
      },
      {
        "slug": "preclinical",
        "label": "Preclinical",
        "label_zh": "临床前",
        "count": 3
      },
      {
        "slug": "proteomics",
        "label": "Proteomics",
        "label_zh": "蛋白质组学",
        "count": 3
      },
      {
        "slug": "state",
        "label": "State model",
        "label_zh": "State 模型",
        "count": 3
      },
      {
        "slug": "tahoe",
        "label": "Tahoe",
        "label_zh": "Tahoe",
        "count": 3
      },
      {
        "slug": "therapy",
        "label": "Therapy",
        "label_zh": "治疗",
        "count": 3
      },
      {
        "slug": "world-model",
        "label": "World model",
        "label_zh": "世界模型",
        "count": 3
      },
      {
        "slug": "zero-shot",
        "label": "Zero-shot",
        "label_zh": "零样本",
        "count": 3
      },
      {
        "slug": "drug-discovery",
        "label": "Drug discovery",
        "label_zh": "药物发现",
        "count": 2
      },
      {
        "slug": "idh",
        "label": "IDH",
        "label_zh": "IDH 突变",
        "count": 2
      },
      {
        "slug": "immune",
        "label": "Immune",
        "label_zh": "免疫",
        "count": 2
      },
      {
        "slug": "interpretability",
        "label": "Interpretability",
        "label_zh": "可解释性",
        "count": 2
      },
      {
        "slug": "multi-omics",
        "label": "Multi-omics",
        "label_zh": "多组学",
        "count": 2
      },
      {
        "slug": "neuro-oncology",
        "label": "Neuro-oncology",
        "label_zh": "神经肿瘤学",
        "count": 2
      },
      {
        "slug": "radiotherapy",
        "label": "Radiotherapy",
        "label_zh": "放射治疗",
        "count": 2
      },
      {
        "slug": "recursion",
        "label": "Recursion",
        "label_zh": "Recursion",
        "count": 2
      },
      {
        "slug": "scaling",
        "label": "Scaling",
        "label_zh": "规模化",
        "count": 2
      },
      {
        "slug": "simulation",
        "label": "Simulation",
        "label_zh": "模拟仿真",
        "count": 2
      }
    ]
  },
  "topic": {
    "description": "Research-graph record, see /data/graph/*.json",
    "fields": {
      "id": "string",
      "type": "enum — cell_state | transition | model | dataset | perturbation | validation | topic",
      "url": "string — path",
      "title_en": "string",
      "title_zh": "string",
      "questions": "{en,zh}[]",
      "entities": "string[]",
      "evidenceRefs": "string[] — item ids",
      "evidence": "{primary[], models[], perturbation_validation[]}",
      "relatedTopics": "string[] — topic ids",
      "summary": "{en,zh}",
      "current_understanding": "{en,zh,refs[]}[]",
      "uncertainties": "{en,zh,refs[]}[]",
      "relations": "relation[] — typed relations scoped to this topic (see relation)",
      "reviewStatus": "string",
      "publicationStatus": "string",
      "lastReviewed": "string|null",
      "version": "string",
      "analysis_files": "string[] — published analysis outputs only"
    }
  },
  "entity": {
    "description": "First-class entity, source records in content/entities/*.json",
    "fields": {
      "id": "string — slug (= file name, = URL segment /entities/<slug>/)",
      "type": "enum — cell_state | gene | pathway | method | drug | model | dataset | disease",
      "name_en": "string — required",
      "name_zh": "string — required",
      "aliases": "string[] — English/Latin aliases",
      "description": "{en,zh} — conservative; supported by evidenceRefs",
      "externalIds": "object — only when verified: hgnc_id, hgnc_symbol, uniprot, cl, mondo, geo",
      "modelKind": "string — optional, \"experimental\" for non-computational model systems",
      "graphNames": "string[] — spellings used in topic `entities` arrays",
      "matchTerms": "string[] — patterns used to derive mentions (\"re:\" case-sensitive regex, \"ri:\" case-insensitive regex, otherwise case-insensitive literal)",
      "mentions": "{items: VC[], topics: topic-id[]} — DERIVED from data/items.csv and content/graph (stale values fail the build; refresh with node scripts/derive-mentions.mjs)",
      "evidenceRefs": "string[] — VC ids supporting the description",
      "relations": "relation[] — outgoing relations stored on the entity record",
      "reviewStatus": "string",
      "version": "string"
    },
    "types": {
      "cell_state": "Recurrent transcriptional cell states described in glioma single-cell studies.",
      "gene": "Genes and mutations that appear in the library’s cell-state and therapy studies.",
      "pathway": "Transcriptional programs, signalling pathways and microenvironment conditions.",
      "method": "Assays, computational approaches, evaluation concepts and treatment modalities.",
      "drug": "Therapeutic agents that appear in library studies.",
      "model": "Computational models and experimental model systems.",
      "dataset": "Atlases, perturbation resources and data platforms.",
      "disease": "Tumor types covered by the library."
    }
  },
  "relation": {
    "description": "Typed relation between two entities, stored in the `relations` array of content/graph/*.json (topic scope) and content/entities/*.json (entity scope); identical subject|predicate|object triples are merged, keeping the lowest confidence",
    "fields": {
      "subject": "entity id",
      "predicate": "enum — promotes | suppresses | marks | transitions_to | validated_by | benchmarked_on | uses_dataset | perturbs | associated_with",
      "object": "entity id",
      "evidenceRefs": "string[] — VC ids (required, non-empty)",
      "evidenceRoles": "object — VC id → role (supports | contradicts | review | dataset | benchmark), required for every evidence item",
      "confidence": "enum — high | medium | low",
      "note": "{en,zh} — required unless confidence is high"
    },
    "predicates": {
      "promotes": "subject promotes or drives object (e.g. a regulator and a cell state)",
      "suppresses": "subject suppresses, represses or restrains object",
      "marks": "subject is a marker or signature for object",
      "transitions_to": "cells in subject state move toward object state",
      "validated_by": "subject finding is tested in object (model system, cohort or experiment)",
      "benchmarked_on": "subject is evaluated against object (benchmark, baseline or challenge)",
      "uses_dataset": "subject is built from, trained on or calibrated against object dataset",
      "perturbs": "subject is a perturbation (assay, drug, treatment) applied to object system",
      "associated_with": "reported association without a stated causal direction"
    },
    "roles": {
      "supports": "item reports or supports the relation",
      "contradicts": "item points the other way or creates tension with the relation",
      "review": "review/perspective item discussing the relation",
      "dataset": "item describes the dataset/resource involved",
      "benchmark": "item is a benchmark/evaluation relevant to the relation"
    },
    "confidence": {
      "high": "stated directly by a cited item (peer-reviewed) or topic text",
      "medium": "stated directly but from a preprint, a model system or partial/observational evidence",
      "low": "inferred, indirect or in tension with other evidence — flagged for editorial review"
    }
  },
  "graph": {
    "node_kinds": {
      "item": "library item (id VCxxxx)",
      "topic": "research topic (id as in content/graph)",
      "hub": "research hub (id hub:<type>)",
      "tag": "facet tag (id tag:<slug>)",
      "entity": "entity (id entity:<slug>)"
    },
    "edge_kinds": {
      "cites": "topic → item: item listed in the topic evidenceRefs",
      "evidence": "topic → item: item placed in the topic evidence map (role = primary | models | perturbation_validation)",
      "uncertainty": "topic → item: item cited in the topic \"what remains uncertain\" section",
      "relatedTopic": "topic → topic: declared related topic",
      "inHub": "topic → hub: topic belongs to a research hub",
      "hasTag": "item → tag: normalised tag (only tags used ≥2 times)",
      "mentions": "entity → item: item title/tags/summary matches the entity (derived from data)",
      "describedBy": "entity → item: item cited as evidence for the entity description",
      "hasEntity": "topic → entity: entity mentioned on the topic page",
      "relation": "entity → entity: typed relation with predicate (promotes | suppresses | marks | transitions_to | validated_by | benchmarked_on | uses_dataset | perturbs | associated_with), confidence (high | medium | low) and evidence items with roles (supports | contradicts | review | dataset | benchmark)"
    }
  }
}