A curated, evidence-backed knowledge base of disease pathophysiology, phenotypes, and treatments linked to authoritative ontologies.
Each disorder is stored as a human-readable YAML file with pathophysiology, phenotypes, genetic factors, and treatments.
All entities are mapped to authoritative ontologies: HPO for phenotypes, MAXO for treatments, GO for processes, and more.
Every claim is backed by PubMed citations with exact quotes, verified against abstracts to prevent misinformation.
AI agents help curate and expand the knowledge base, with automated validation to catch errors and hallucinations.
LinkML schema ensures data integrity with automated checks for structure, term validity, and reference accuracy.
Fully open source with version-controlled YAML files. Contribute via pull requests or use the data in your own projects.
Every entity in dismech is mapped to authoritative biomedical ontologies, enabling interoperability and semantic precision.
Search across 55 disorders with faceted filtering by phenotype, treatment, cell type, and more.
Launch Browser