概要
Classify drugs with confidence
As new drugs outpace the creation of reference spectra, Wiley Drug Classification Models classify an unknown spectrum into a drug class, like opioids, fentanyls, or benzodiazepines, even when no matching reference spectrum is available.
Current techniques:
- IR
- Raman
- GC-MS
Available as a standalone API, or use the models directly within KnowItAll.
Features
- Returns a predicted drug class and confidence score per call
- Individual or batch classification
- Built on Wiley's authoritative spectral data
- 5-fold cross-validated models to ensure result reliability and consistency
- Model version included with every result
- Optional request ID for easier traceability and auditing
- Inference-only, secure data handling: Your data is never stored or used to train model
For models used within KnowItAll, you can automate classifications using the KnowItAll Automate SDK if your workflow requires it
Applications
Use this API to:
- Submit a spectrum and receive a predicted drug class
- Classify spectra you can't identify with spectral libraries alone
- Validate or confirm results from your spectral searches
- Make quick first-pass classifications
Built for:
- Forensic analysis and toxicology laboratories
- Law enforcement and homeland security
- Drug analysis laboratories
- Instrument partners
- ELN, CDS, and LIMS providers
Available Models
- IR models 99% accuracy · 77% precision · 88% recall
- Natural opioids
- General opioids
- Fentanyls and related compounds
- Steroids and related compounds
- Ketamine and related compounds
- Cathinones and related compounds
- GC-MS models 99% accuracy · 84% precision · 93% recall
- Natural opioids
- Synthetic opioids
- General opioids
- Fentanyls and related compounds
- Piperazines
- Benzodiazepines
- Steroids
- Ketamine and related compounds
- Cocaine and related compounds
- Cathinones and related compounds
- Raman models 99% accuracy · 88% precision · 99% recall
- Steroids