MAKO: schema, aggregate statistics, and ontology mapping rules for a multimodal Alzheimer's-disease knowledge graph

Derived, non-identifying artefacts from MAKO (Multimodal Alzheimer's Knowledge graph with Ontology grounding), a Neo4j knowledge graph built on the ADNI cohort. This deposit contains: the graph SCHEMA (node labels, relationship types, and the OntologyConcept inventory across eight source ontologies — SNOMED-CT, LOINC, UBERON, HPO, ICD-10, MONDO, DOID, and the Gene Ontology); pre-computed AGGREGATE STATISTICS (node/edge counts, URI and ontology coverage, the FAIR scorecard, the structural-validity rubric, cross-vocabulary alignment to AlzKB, and the per-step audit); and the curated column-to-concept MAPPING RULES, together with a data dictionary and pointers to the source code. No participant-level ADNI data is included. Under the LONI Image and Data Archive Data Use Agreement, ADNI participant data cannot be redistributed; the underlying graph is not deposited. Every file was scanned for participant identifiers before inclusion. Source code: https://github.com/theFellandes/ADNIKnowledgeGraph