OncoQuery Assistant

AI-powered genomic variant interpretation tool for oncologists, researchers, and students. Rapidly interpret genomic variants by querying the CIViC database and generating comprehensive clinical reports using Google Gemini AI.

🌐 Live Demo: https://onco-query-assistant.vercel.app/

Features

• Variant Input: Enter gene names and variants manually (supports HGVS and CIViC formats)
• CIViC Integration: Direct GraphQL queries to the CIViC database for clinical evidence
• AI Report Generation: Generate structured clinical reports with:
  • Variant Significance
  • Clinical Relevance (Diagnostic, Prognostic, Predictive)
  • Treatment Options (FDA-approved, NCCN Category 1, investigational)
  • Evidence Summary
• Interactive Chat: Ask follow-up questions about variants, treatments, and terminology

Tech Stack

• Frontend: Next.js 14+ with App Router, TypeScript, Tailwind CSS
• Backend: Next.js API Routes
• AI: Google Gemini 1.5 Pro (reports) and Flash (chat)
• Data Source: CIViC Database (GraphQL API)

Getting Started

Prerequisites

• Node.js 18+ and npm
• Google Gemini API key (from GCP)

Installation

1. Clone the repository:

git clone https://github.com/werther41/onco-query-assistant.git
cd onco-query-assistant

2. Install dependencies:

npm install

3. Create .env.local file in the root directory:

GOOGLE_GEMINI_API_KEY=your_gemini_api_key_here
CIVIC_GRAPHQL_URL=https://civicdb.org/api/graphql
NEXT_PUBLIC_APP_URL=http://localhost:3000

4. Run the development server:

npm run dev

5. Open http://localhost:3000 in your browser.

Usage

1. Enter Variant Information:

   • Enter a gene name (required)
   • Optionally enter variant information (e.g., T790M, G12S, or p.Arg361Cys)
   • The system automatically normalizes HGVS format to CIViC format

2. Generate Report:

   • Click "Generate Report"
   • The system queries CIViC and generates an AI-powered clinical report

3. Ask Questions:

   • After viewing the report, use the chat interface to ask follow-up questions
   • Examples: "What is a kinase inhibitor?", "Explain evidence level A"

Example Queries

• EGFR - T790M
• MET - Exon 14 Skipping
• KRAS - G12S (or p.Gly12Ser)
• TP53 - R248W (or p.Arg248Trp)

Project Structure

oncoQuery/
├── src/
│   ├── app/              # Next.js App Router
│   │   ├── api/          # API routes
│   │   ├── report/       # Report display page
│   │   └── page.tsx      # Home page
│   ├── components/       # React components
│   ├── lib/              # Core logic
│   │   ├── civic/        # CIViC integration
│   │   └── gemini/       # Gemini AI integration
│   └── types/            # TypeScript types
└── content-source-reference/  # Reference documentation

Development

• npm run dev - Start development server
• npm run build - Build for production
• npm run start - Start production server
• npm run lint - Run ESLint

Notes

• Reports are stored in browser sessionStorage (not persisted to database)
• Chat history is maintained in component state during the session
• The app uses direct GraphQL queries to CIViC (not the MCP server)

License

Private project - All rights reserved