It works in the average patient — trials are not average patients
Without a separate pharmacogenomics database subscription or a clinical bioinformatics team. CYP metabolism profile, PGx population breakdown, and resistance variant flags — in seconds, not weeks.
“Which patient populations will respond to this EGFR inhibitor given known CYP2D6 polymorphisms and resistance variants?”
How it works
Submit a candidate and target gene
Provide the SMILES, target gene symbol, and optional ADMET results. The platform cross-references 56 pharmacogene profiles, 135K ClinVar pathogenic variants, and CYP metabolism from upstream ADMET predictions.
Population-level analysis
CYP2D6, CYP2C9, CYP3A4 metabolizer phenotypes by population. Resistance variants affecting the binding site flagged. HGNC gene symbol validation against 44K symbols + 58K aliases.
Clinical viability summary
Which populations will respond, which will metabolize too fast or too slow, where resistance is prevalent. A clear stratification output for clinical planning — not a raw data dump.
Proof
56 pharmacogene documents in omics_pgx (Cosmos DB). 134,940 ClinVar pathogenic variants in omics_resistance. 13,252 with affects_binding_site = true.
HGNC gene symbol validation: 44K symbols + 58K aliases. CYP substrate analysis from ADMET results (CYP3A4, CYP2D6, CYP2C9 substrate probabilities).
Direct Cosmos DB reads — no backend service dependency. Results in seconds for any target gene.
Know which patients will respond
56 pharmacogenes. 135K variants. Sign up and stratify your first candidate.