PolyRisk

PolyRisk is a web app (with optional desktop CLI) that helps genetics teams compute, validate, and report polygenic risk scores (PRS) from genotype data without building a brittle in-house pipeline. Users upload PLINK/VCF data (or point to cloud storage), select a published GWAS/score file, and the app runs standardized QC, ancestry-aware scoring, and calibration checks. It outputs a clean, audit-friendly report: score distributions, performance metrics (AUC/R2 where labels exist), population stratification warnings, and a reproducible run manifest. This is not a consumer health app; it’s aimed at research and clinical-lab-adjacent teams who need repeatable PRS workflows and documentation. The app is a combination traditional + AI: traditional for the pipeline, AI for automated “what went wrong” diagnostics, parameter suggestions, and plain-language summaries for internal stakeholders. Brutal truth: PRS is messy and often overpromised—this product wins by being conservative, transparent, and reproducible.