PredomicsApp Web

Modern web application for gpredomics — sparse, interpretable ML model discovery for omics data.

Architecture: FastAPI (Python) backend + Vue.js frontend, powered by gpredomicspy.

Quick Start

Backend (development)

cd backend
python -m venv .venv
source .venv/bin/activate
pip install -e ".[dev]"

# Optional: install gpredomicspy for real ML engine
# pip install ../gpredomicspy  # or: cd ../gpredomicspy && maturin develop

uvicorn app.main:app --reload

API docs at http://localhost:8000/docs

Frontend (development)

cd frontend
npm install
npm run dev

Frontend at http://localhost:5173

Docker (production)

docker compose up --build

App at http://localhost:8000

API Endpoints

Method	Endpoint	Description
GET	/health	Health check + engine status
POST	/api/projects/	Create project
GET	/api/projects/	List projects
GET	/api/projects/{id}	Get project
DELETE	/api/projects/{id}	Delete project
POST	/api/projects/{id}/datasets	Upload dataset
POST	/api/analysis/{id}/run	Launch analysis
GET	/api/analysis/{id}/jobs/{jid}	Job status
GET	/api/analysis/{id}/jobs/{jid}/detail	Full results

Project Structure

predomicsapp-web/
├── backend/
│   ├── app/
│   │   ├── main.py          # FastAPI app
│   │   ├── core/config.py   # Settings
│   │   ├── models/schemas.py # Pydantic models
│   │   ├── routers/         # API endpoints
│   │   └── services/        # Business logic
│   └── tests/
├── frontend/                # Vue.js SPA
├── data/qin2014_cirrhosis/  # Demo dataset (Qin et al. 2014)
├── Dockerfile
└── docker-compose.yml

Tests

Running Tests

cd backend
pip install -e ".[dev]"
pytest -v

Coverage Summary

The backend test suite contains 20 tests organized in 5 categories, covering all API endpoints and input validation:

Test Class	Tests	What's Covered
TestHealth	2	/health endpoint returns correct fields (status, engine_available, version); status is always "ok"
TestProjects	7	Full CRUD lifecycle — create, list, get, delete; 404 on missing project; unique IDs; ISO timestamps
TestDatasets	4	TSV and CSV file upload via multipart; 404 on missing project; project metadata updated after upload
TestAnalysis	3	/run returns job_id with pending status; 404 on missing project; empty job list for new projects
TestSchemaValidation	4	Rejects invalid algo values (422); rejects invalid fit values (422); accepts valid non-default params; defaults are applied correctly

Total: 20 tests — all passing.

What's Tested

• All 9 REST API endpoints (see table above)
• Request validation via Pydantic schemas (invalid enum values return 422)
• File upload with correct content-type handling
• Project isolation (operations on non-existent projects return 404)
• Default parameter injection for analysis configuration
• Async request handling (tests use httpx.AsyncClient with ASGITransport)

Not Yet Covered

• End-to-end analysis execution (requires gpredomicspy + sample data)
• WebSocket/SSE progress streaming (not yet implemented)
• Frontend component tests (planned for Phase 3)
• Docker build integration tests

CI/CD

GitHub Actions runs on every push and PR:

• backend-test: Installs dependencies and runs pytest
• frontend-build: Installs npm packages and runs npm run build
• docker-build: Builds the multi-stage Docker image

See .github/workflows/ci.yml.

License

GPL-3.0