/bio/informat >projects
← back to vcfclick

vcfclick hosted preview

Explore a VCF cohort without downloading the VCF

Ask a genomics question, inspect the generated SQL, and query a 1000 Genomes phase 3 cohort directly in your browser. This public demo uses DuckDB-Wasm over Parquet; vcfclick also supports embedded chDB for ClickHouse-backed cohort databases with no ClickHouse server to operate.

Source: github.com/nuin/vcfclick · PyPI

2,504
samples
~243k
variants
DuckDB-Wasm
browser demo
chDB
ClickHouse backend
Want to query your own cohort like this? This demo runs on public 1000 Genomes data. We're exploring a hosted version for private cohorts — tell us about yours →
Starting…
try:

Hosted workspaces

Want this for your cohort?

Request hosted cohort workspace

Hosted vcfclick workspaces turn VCF-derived cohorts into private, shareable query surfaces. Use DuckDB/Parquet for lightweight browser sharing and chDB-backed workspaces when cohort-scale genotype queries need ClickHouse-style performance.

For labs

  • Private cohort workspace for PI and lab members
  • Saved SQL, visible generated queries, CSV/Parquet export
  • Local-first CLI path when data cannot leave the workstation

For cores

  • Project workspaces instead of static TSV handoffs
  • Shareable collaborator access with auditable SQL
  • Bundle import/export for repeatable delivery workflows

Runs DuckDB-Wasm in your browser. You bring your own API key for either Google Gemini (free tier) or Anthropic Claude — it stays in your browser's localStorage and is only ever sent to that provider's own API endpoint. Parquet files are fetched by HTTP range read, so you only download the columns and row groups each query touches. The full vcfclick CLI can also use embedded chDB, giving labs and cores a ClickHouse-compatible backend without operating a ClickHouse server.