FAQ¶
NosoGraph v0.2.1. Disease Intelligence. Connected. Open-source research software for connecting disease knowledge, evidence, and provenance across biomedical sources.
Who is it for? Researchers, developers, students, curators, and institutions evaluating open research software.
Is NosoGraph medical advice? No.
How many disease modules are in the registry? 10,404 discoverable modules (Disease.list_all()) in the current snapshot; 10,406 on-disk directories including two blocked slugs; 10,391 harvest-catalog ids. Public L2=88 and L3=2 are an n=500 sample. Registry breadth is not "supported diseases." Counts: public-status.yaml and coverage.
Are all diseases equally curated? No. Registry ≠ curation depth.
Where does the data come from? Upstream ontologies and databases. Sources.
Can I use it offline? CLI and fixture-backed tests can run offline. Live connectors need the network.
Does it require an LLM? No. Optional LLM enrichment is experimental.
Does it use OpenAI? Only if you set OPENAI_API_KEY for optional workflows.
Can I add a disease? Yes, via the curation path.
Can I add a data source? Propose it with the data-source issue template; integration is a reviewed engineering change.
Can I use it in academic research? Yes, with citation and license/data-term compliance.
How do I cite it? Citation.
What license? Apache-2.0 for code; upstream terms for data.
Can I deploy it myself? Yes. See self-hosted deployment.
What is NosoGraph Compare? A beta workflow for deterministic two-to-five-disease comparisons with explicit missingness, evidence drill-down, and JSON/Markdown exports.
What is the Evidence Workspace? A BETA workflow that assembles evidence into claims and ranked hypotheses.
Why is the Python package still called med-research? Compatibility. See package naming.