chemgraphbuilder 0.1.23

A package to build chemical knowledge graphs using data from PubChem and Neo4j

ChemGraphBuilder

chemgraphbuilder is a Python package designed for transforming chemical data into knowledge graphs.
It leverages PubChem for data extraction and Neo4j for building graph databases, enabling researchers to efficiently extract, process, and visualize complex chemical relationships with precision.
The package is designed for easy extension to include other data sources in future releases.

---

Table of Contents

• ChemGraphBuilder
  • Table of Contents
  • Neo4j Requirements
  • 🚀 Quick Start
    • 1️⃣ Install and Run Neo4j
    • 2️⃣ Install chemgraphbuilder
    • 3️⃣ Connect to Neo4j in Python
    • 4️⃣ First Example
  • Usage
    • From Python
    • From Command Line
  • Features
  • Documentation
  • Contributing
  • License
  • Contact
  • Acknowledgments

---

Neo4j Requirements

chemgraphbuilder requires a running Neo4j database that is accessible via Bolt URI, username, and password.

You can run Neo4j:

• Locally (Neo4j Desktop or Docker)
• Remotely (Neo4j Aura Cloud)

Default Bolt port: 7687
Default Web UI port: 7474

---

🚀 Quick Start

Follow these steps to get up and running with chemgraphbuilder and Neo4j in under 5 minutes.

1️⃣ Install and Run Neo4j

Option A – Docker (fastest)

docker run \
  --name neo4j \
  -p 7474:7474 -p 7687:7687 \
  -e NEO4J_AUTH=neo4j/testpassword \
  neo4j:5.14

• Bolt URI: bolt://localhost:7687
• Username: neo4j
• Password: testpassword
• Web UI: http://localhost:7474

Option B – Neo4j Desktop

1. Download from: https://neo4j.com/download/
2. Create a new project and database.
3. Note the Bolt URI, username, and password.

---

2️⃣ Install chemgraphbuilder

pip install chemgraphbuilder

Or visit the PyPI Project Page for the latest release.

---

3️⃣ Connect to Neo4j in Python

from chemgraphbuilder import Neo4jBase

# Connection details
NEO4J_URI = "bolt://localhost:7687"
NEO4J_USER = "neo4j"
NEO4J_PASSWORD = "testpassword"

# Connect and test
db = Neo4jBase(uri=NEO4J_URI, user=NEO4J_USER, password=NEO4J_PASSWORD)
db.test_connection()

---

4️⃣ First Example

# Create a simple test node
db.run_query("CREATE (:Test {name: 'Hello Neo4j'})")
print("Node created!")

Check in the Neo4j Browser:

MATCH (n) RETURN n;

---

Usage

From Python

from chemgraphbuilder.setup_data_folder import SetupDataFolder
from chemgraphbuilder.node_collector_processor import NodesCollectorProcessor
from chemgraphbuilder.relationship_collector_processor import RelationshipsCollectorProcessor
from chemgraphbuilder.graph_nodes_loader import GraphNodesLoader
from chemgraphbuilder.graph_relationships_loader import GraphRelationshipsLoader

# Setup data folder
setup_folder = SetupDataFolder()
setup_folder.setup()

# Collect nodes
collector = NodesCollectorProcessor(node_type=node_type, enzyme_list=enzyme_list, start_chunk=0)
collector.collect_and_process_data()

# Collect relationships
collector = RelationshipsCollectorProcessor(relationship_type=relationship_type, start_chunk=0)
collector.collect_relationship_data()

# Load nodes into Neo4j
graph_nodes_loader = GraphNodesLoader(uri, username, password)
graph_nodes_loader.load_data_for_node_type(label)
graph_nodes_loader.close()

# Load relationships into Neo4j
graph_relationships_loader = GraphRelationshipsLoader(uri, username, password)
graph_relationships_loader.add_relationships(relationship_type)
graph_relationships_loader.close()

From Command Line

setup-data-folder
collect-process-nodes --node_type Compound --enzyme_list gene1,gene2 --start_chunk 0
collect-process-relationships --relationship_type Assay_Compound --start_chunk 0
load-graph-nodes --uri bolt://localhost:7687 --username neo4j --password password --label Compound
load-graph_relationships --uri bolt://localhost:7687 --username neo4j --password password --relationship_type Assay_Gene

More examples: Usage Examples.

---

Features

• Node Representation: Compounds, genes, proteins, bioassays.
• Comprehensive Relationships: Includes assay-compound, assay-gene, compound similarity, co-occurrence, inhibitor/activator/ligand, etc.
• Data Integration: Schema supports adding new sources.
• Flexible Access: Command line & Python API.

---

Documentation

Full docs: ChemGraphBuilder Documentation

---

Contributing

Issues: GitHub Issues Pull requests welcome.

---

License

GPL-3.0 – see LICENSE.

---

Contact

Asmaa A. Abdelwahab – [e-mail]

---

Acknowledgments

• PubChem – for chemical and bioassay data.
• Neo4j – for graph database capabilities.