Merged PGxAI + antidepressant expert-system prototype
PGxAI Clinical Engine v2 — Evidence-aware pharmacogenomic antidepressant ranking with safety gates
A polished educational clinical decision-support prototype showing how bioinformatics, deterministic rules, and product design can make pharmacogenomic antidepressant review more transparent.
Safety disclaimer
PGxAI Clinical Engine v2 is an educational clinical decision-support prototype, not medical advice. It does not diagnose, prescribe, guarantee response, or replace clinician judgment. Outputs describe relative PGx compatibility in this prototype and require professional interpretation before any clinical action.
Genotype-to-phenotype interpretation
Built for a 30-second understanding: recommendation bucket, why, rules, evidence category, and missing data.
Explainable ranking
Built for a 30-second understanding: recommendation bucket, why, rules, evidence category, and missing data.
Safety-first workflow
Built for a 30-second understanding: recommendation bucket, why, rules, evidence category, and missing data.
Hackathon-ready demo
Built for a 30-second understanding: recommendation bucket, why, rules, evidence category, and missing data.
What PGxAI demonstrates
PGxAI converts CYP2D6/CYP2C19 star-allele genotype inputs into metabolizer phenotypes, then combines that interpretation with clinical context in a transparent, deterministic medication-safety report.
- • Simple and Detailed Search modes starting from CYP2D6/CYP2C19 diplotypes.
- • Local genotype-to-phenotype mapping followed by layered PGx, phenoconversion, safety, preference, and missing-data rules.
- • Ranked antidepressants with dimensional scoring rather than a single unexplained score.
- • Export-to-JSON and print-friendly report actions.
Why this matters
Pharmacogenomic results can be hard to interpret during routine care. A transparent assistant can organize risk signals while keeping clinician judgment central.
This MVP is intentionally not a certified medical device. It shows how governed rule ingestion, evidence categories, auditability, and safety-first UX could create value in biomedical sciences.
Future roadmap preview
Production versions would add rule versioning, clinical governance, validated source ingestion, FHIR/CDS Hooks, and audit logs.