This handles clinical interpretation of structural variants like deletions, duplications, and inversions using ACMG-adapted criteria. It walks through a seven-phase workflow: normalizing coordinates, identifying affected genes, pulling ClinGen haploinsufficiency and triplosensitivity scores, checking gnomAD population frequencies, and scoring pathogenicity on a 0-10 scale before mapping to the five-tier clinical classification. The dosage sensitivity logic is explicit: a deletion is only pathogenic if it removes a haploinsufficient gene, a duplication matters when the gene is triplosensitive. It enforces a compute-first approach, running Python for scoring rather than describing what you'd do. Use this when you need structured evidence grading for CNV or chromosomal rearrangement reports.
npx skills add https://github.com/mims-harvard/tooluniverse --skill tooluniverse-structural-variant-analysis