Predict · Module 01
Ancestry-informed polygenic risk scoring.
650+ validated polygenic models, 200M+ variants imputed at 99.9% accuracy, validated across 1M+ patients.
Ancestry-informed polygenic risk scoring at clinical scale.
650+ validated polygenic models
A continuously expanding catalog of polygenic risk scores spanning the most prevalent and consequential conditions in oncology, cardiology, metabolic, and neurological medicine.
Multi-ancestry calibration
Polygenic risk scores calibrated across diverse ancestries so reporting performs reliably for admixed and non-European patients.
200M+ variants per sample
Genotype data expanded to over 200 million imputed variants at 99.9% accuracy, enabling near-whole-genome predictive density from genotyping-array inputs.
Ensemble methodology
Top-performing polygenic methods combined through an ensemble approach, with disease prediction enhanced by clinical and demographic context.
Validated in peer-reviewed research
Methods published in Nature Communications (2025), Scientific Reports (2025), and medRxiv (2023), describing ancestry inference, imputation, and disease-prediction models.
Clinical benchmark performance
Coronary-artery-disease polygenic models identify 55–80× more true coronary events than rare-pathogenic-variant testing alone; 12 of 30 disease models exceed 80% AUC across validation cohorts.
Predict module specifications.
Performance figures from Optimization of multi-ancestry polygenic risk score disease prediction models, Scientific Reports, May 2025. Imputation methodology from Empowering GWAS discovery through enhanced genotype imputation, medRxiv, Dec 2023.