Input boundary
- One required de-identified case description
- Optional filtered genomic text or phenotype file
- No names, dates of birth, record numbers, or addresses
METHODS BEFORE ANSWERS
Useful uncertainty.
Visible limitations.
RDK is a research prioritization tool. It combines case inputs with current source retrieval, but it does not establish causality, diagnose a condition, or recommend treatment.
Run an authorized analysis โ
Methodology
Model task
- Rank plausible causal variants
- Explain phenotype and mechanism fit
- Generate exploratory, mechanism-linked repurposing directions
- Find current primary or authoritative sources
Structured response
The API is constrained to a defined schema so the interface can show each field directly rather than guessing how to interpret free-form prose.
Traceability
The site exposes the model version, response ID, token usage, source list, limitation list, and every evidence statement returned by the run.
Safety and interpretation
Not a diagnosis
A high model score is a prioritization signal, not proof of pathogenicity or disease causation.
Not a prescription
Drug directions are research hypotheses only. They require mechanism replication, dose plausibility, interaction review, and expert oversight.
Human review required
Confirm build, transcript, allele phase, inheritance model, read-level support, population frequency, and phenotype fit using validated tools and qualified experts.
Source quality varies
Open every cited source. Confirm that it supports the exact claim, applies to the relevant organism and disease context, and has not been overgeneralized.
Privacy boundary. The OpenAI key stays on the server. Case excerpts are transmitted only after sign-in and explicit confirmation. Researchers remain responsible for authorization and de-identification.