What problem does it solve?
This skill automates the calculation and interpretation of polygenic risk scores (PRS) from direct-to-consumer genotype files, removing manual matching, weighting, and statistical conversion steps so users can quickly obtain trait-level risk estimates with provenance and reproducibility.
Core Features & Use Cases
- PGS Catalog integration: Search the PGS Catalog, download harmonised scoring files, and use curated offline scores for fast demo runs.
- Genotype parsing & matching: Parse 23andMe and AncestryDNA formats (including gzipped files), match rsIDs to scoring files, and compute dosage-weighted PRS.
- Percentile estimation & safety: Estimate percentiles using curated reference mean/SD or allele-frequency-based variance, annotate risk categories, warn on low coverage or ancestry mismatch, and always keep genotype data local.
- Reporting & reproducibility: Emit a structured report directory with a Markdown narrative, CSV score table, optional figures, commands.sh, environment.yml, and checksums for reproducible review.
- Use cases: Individual risk checks for type 2 diabetes or BMI from a 23andMe file, batch scoring of curated traits for research demos, and integration with ancestry or pharmacogenomics skills for contextualised genomic reports.
Quick Start
Calculate my Type 2 diabetes PRS from this 23andMe file and save a Markdown report, CSV scores, and optional figures to the specified output directory.