Installation

This guide covers setting up the MCBO development environment.

Prerequisites

• Conda or Miniconda

Conda is used in make conda-env to install:

• Python 3.9+ (3.10 recommended)

• Java (for ROBOT ontology tool)

• robot

Creating the Conda Environment

1. Create a new conda environment named mcbo:

make conda-env

2. Activate the environment:

conda activate mcbo

Installing the MCBO Package

Install the mcbo Python package and ROBOT:

# Using Make (recommended) - installs Python packages + ROBOT
make install

This provides the following CLI commands:

• mcbo-csv-to-rdf - Convert CSV metadata to RDF instances

• mcbo-build-graph - Build graphs from study directories

• mcbo-run-eval - Run SPARQL competency queries

• mcbo-stats - Generate graph statistics

ROBOT

ROBOT is automatically downloaded by make install. It’s used for ontology quality control checks and is located at .robot/robot.jar.

To verify ROBOT works:

java -jar .robot/robot.jar --version

To re-download manually:

make robot

Verifying Installation

Run the demo workflow to verify everything is set up correctly:

# Activate environment
conda activate mcbo

# Run the full demo
make demo

This will:

1. Build the demo graph from data.sample/

2. Run all 8 competency question queries

3. Display statistics and results

You should see output ending with:

✅ Demo data processing complete
   Graph: data.sample/graph.ttl
   Results: data.sample/results/

Dependencies

Core Python Dependencies

• rdflib>=7.0.0 - RDF graph manipulation

• pandas>=1.5.0 - Data processing

External Tools

• Java - Required for ROBOT (JRE 8+)

• ROBOT - Ontology tool for QC queries (located at .robot/robot.jar)

See requirements.txt for the complete list of Python dependencies.