Loading...
Disclaimer: Informational only — not medical advice. AI hypotheses are not clinically validated. Medic Oath is for informational and educational purposes only — not medical advice. AI-generated hypotheses have not been clinically validated. Always consult a qualified healthcare professional. Privacy · Terms
One page. Two engines. From disease to treatment hypothesis in under 60 seconds.
What it does: Generates 10-20 scored drug repurposing hypotheses for any disease using dual-model AI consensus (Claude Opus + Sonnet) across 96+ biomedical data sources.
When to use: First step for any disease. Quickly generates many hypotheses to explore. Best for discovering new treatment angles.
Output: Scored hypothesis cards with confidence, novelty, impact, feasibility
Includes: Biomarkers, safety data, animal model evidence, cheapest validation experiment, special population notes
Export: PDF report + CSV data
Enrichment: 28-feature computational verification (pathway, ADMET, DDI, prior art, compounds, FAERS signals)
How to use
What it does: Takes a specific hypothesis from the Hypothesis Engine and runs 8 specialist AI agents against it — safety gate, bioinformatics, drug repurposing, COI scan, experiment design, and ethics review.
When to use: After generating hypotheses — click "Validate with Prometheus" on your most promising hypothesis to get safety assessment, trial design, and ethics review before advancing to pre-clinical.
Output: Comprehensive research report with expandable agent sections
Includes: Safety gate (LD50/NOAEL), ethics review (Belmont), trial design (sample size, endpoints), COI scan
Agents: 8 specialist agents with real-time progress visualization
Pipeline stage: Pre-clinical and clinical trial preparation
How to use
Both tools work together in the path to a cure. Here's when and why you'd use each.
| Feature | Hypothesis Engine | Prometheus |
|---|---|---|
| Purpose | Generate many hypotheses quickly | Deep-validate one specific hypothesis |
| Input | A disease name | A hypothesis from the Hypothesis Engine |
| Output | 10-20 scored hypothesis cards | GO / CAUTION / NO-GO verdict + full report |
| AI Models | Claude Opus + Sonnet (consensus) | Claude Sonnet (8 sequential agents) |
| Data Sources | 96+ biomedical databases | PubMed + targeted literature for the hypothesis |
| Safety | ADMET + DDI flagging | Full safety gate (LD50, NOAEL, black box, DDI, contraindications) |
| Ethics | Special populations noted | Belmont Report screening + vulnerable population analysis |
| Trial Design | Not included | Phase, sample size, endpoints, cost, regulatory pathway |
| COI Check | Not included | Flags industry-funded citations and author conflicts |
| Best For | Exploring treatment options | Deciding whether to advance to pre-clinical |
MedicOath — Finding cures the world forgot to look for.
AGPL-3.0 · CC0-1.0 · medicoath.com