| name | clinical-variant-agent |
| description | Full clinical variant workup — gnomAD population frequency, ClinVar significance, protein domain impact, AlphaFold structure context, and PubMed literature |
Perform a comprehensive clinical variant workup for: $ARGUMENTS
Use the MCP tools to gather data from all relevant sources, then synthesize a structured clinical variant report. Follow the steps below in order. If a step fails or returns no data, note the gap and continue.
Input Parsing
Determine the variant format:
- rsID (e.g.,
rs11549407) — use directly for gnomAD lookup
- chrom-pos-ref-alt (e.g.,
7-140753336-A-T) — use directly for gnomAD
- HGVS notation (e.g.,
NM_007294.4:c.5266dupC) — search ClinVar first to get genomic coordinates
- Gene + protein change (e.g.,
BRAF V600E) — search ClinVar first, extract variant details
Data Gathering Steps
1. Population Frequency (gnomAD)
- Call
gnomad_get_variant with the variant ID or rsID.
- Extract: overall allele frequency (exome + genome), per-population frequencies, homozygote counts, filter flags.
- If the variant is not found in gnomAD, note this — absence from gnomAD is itself informative (suggests very rare).
2. Clinical Significance (ClinVar)
- Call
clinvar_search with the variant identifier (rsID, HGVS, or gene + variant description).
- Extract: clinical significance classification, review status (star rating), associated conditions/diseases, submitter count.
- Note any conflicting interpretations.
3. Gene Context
- From the gnomAD or ClinVar result, identify the affected gene symbol.
