Docker Guide

PBI-Scope uses Docker Compose to orchestrate three services: pipeline, analysis, and api.

Quick Start

# Clone and configure
git clone https://github.com/ThibaultSchowing/PBI.git
cd PBI
cp .env.example .env
# Edit .env: set NCBI_EMAIL, UID, GID

# Build and run pipeline
docker compose build pipeline
docker compose run --rm pipeline

# Start analysis container
docker compose build analysis
docker compose up -d analysis

See the Installation Guide for detailed setup instructions.

Services

Service	Purpose	Port
pipeline	Builds/updates the database	—
analysis	Jupyter Lab + VS Code Dev Containers	8889
api	REST API for metadata queries	8000

Security

The analysis container runs Jupyter Lab with authentication disabled for local development. See Analysis Container Guide for details and hardening steps.

UID/GID Setup

Both pipeline and analysis services run as your host user to prevent root-owned files. Set UID and GID in .env after cloning:

echo "UID=$(id -u)" >> .env
echo "GID=$(id -g)" >> .env

See Installation Guide for full explanation.

API Service

The API provides metadata queries and SQL exploration without the full pbi package:

docker compose up api
# Available at http://localhost:8000

See API Reference for endpoint documentation.