Q Q-omics
Request access

Cancer multi-omics, pre-validated before you ever query it.

Q-omics runs statistics across hundreds of sampling splits before a result ever reaches you — so a hit means something held up, not that it happened to clear p<0.05 once. Query it as an AI-agent tool today, or call the REST API directly once general access opens.

38KGenes
34+20+10Tissue / cell / CPTAC lineages
561Drug compounds
2Consensus axes
POST /qomics/api/plot
{
  "class_name": "box_plot",
  "q_data": {
    "sample_type": "Tissue",
    "lin_code": "BRCA",
    "x_type": "rna",
    "x_symbol": "TP53"
  }
}

// 200 OK
{
  "chart_type": "box",
  "stats": {
    "p_value": 3.1e-08,
    "fold_change": -1.842
  },
  "criteria": "Comparison: Group1 (Low) vs
    Group2 (High) rna — median split;
    Welch t-test",
  "data": { "counts": { "group1": 274, "group2": 273 } }
}

Two ways in

Same curated warehouse, two doors — pick by what you're building.

Live — preview

Building an AI agent or copilot?

Connect Claude, ChatGPT, or any MCP-compatible client. Your LLM resolves genes, drugs and lineages in natural language, then calls 17 read-only tools — survival, correlation, association, normal-vs-tumor, and more — across guided analyses and a PQL query layer for anything the fixed tools don't cover.

  • OAuth 2.1 + PKCE, sign in with a Q-omics account
  • 17 tools · 2 toolsets (Guided Analysis + PQL)
  • Every call runs under a locked, read-only query guard
Connect your LLM →
Opening soon

Building a dashboard, pipeline, or internal tool?

Call the same curated data directly over JSON — browse hit tables, pull summary rollups, render plots, or run the sampling-consensus grid — without going through an LLM. The contract below is stable; general access is opening in phases.

  • 28 endpoints — browse, summary, plot, consensus, subtype
  • One request envelope, four response shapes
  • API-key auth, issued on approval
Read the API spec →

What's inside

Pre-computed and validated across sampling splits and cancer lineages before it's ever queried — not raw counts run fresh per request.

34TCGA tissue lineages · RNA, mutation, RPPA, immune infiltration
10CPTAC lineages · mass-spec proteome + phosphoproteome
20CCLE cell lines · +CRISPR/shRNA dependency, methylation
6+13Pembrolizumab cohorts · NCI60 drug-response lineages

Docs & resources

Everything a client needs to build against — the reference, the data it's built on, and the trust questions that come before either.

Q-omics is in developer preview. The MCP connector is open today; direct REST API access is opening in phases as usage limits and API keys roll out. Request access to get on the list.