Anatomy of a Biomedical Knowledge Graph

Run the Anatomy of a Biomedical Knowledge Graph MicroSim Fullscreen

About This MicroSim

This MicroSim displays a heterogeneous biomedical knowledge graph — a graph where nodes and edges have different types. Students can explore how genes, diseases, drugs, phenotypes, and Gene Ontology (GO) terms are interconnected through typed, labeled relationships.

Node Types

Genes (blue) — Gene/protein entities (e.g., BRCA1, TP53, EGFR)
Diseases (red) — Disease entities (e.g., breast cancer, lung cancer)
Drugs (green) — Pharmaceutical compounds (e.g., tamoxifen, erlotinib)
Phenotypes (orange) — Observable characteristics (e.g., tumor growth, inflammation)
GO Terms (purple) — Gene Ontology functional annotations

Edge Types (Relationships)

gene–disease: "associated_with" — a gene is linked to a disease
drug–gene: "targets" — a drug acts on a gene/protein
drug–disease: "treats" — a drug is used to treat a disease
gene–phenotype: "causes" — a gene mutation causes a phenotype
gene–GO: "annotated_with" — a gene has a functional annotation

Why This Matters

Biomedical knowledge graphs power drug repurposing, disease gene discovery, and clinical decision support. Understanding their structure is essential for bioinformaticians working with graph databases and machine learning on biological knowledge.

How to Use

Click any node to see its details and all connected relationships
Hover over edges to see the relationship type
Drag nodes to explore the graph layout
Zoom to focus on specific regions of the graph
Trace paths — Follow multi-hop paths like Drug → targets → Gene → associated_with → Disease to understand drug mechanisms

Suggested Exploration

Find a drug node and trace its path: which gene does it target, and what diseases is that gene associated with?
Find a gene connected to both a disease and a GO term — the GO term helps explain why the gene is associated with the disease
Look for triangles: Drug → targets → Gene → associated_with → Disease. Does the drug also have a direct "treats" edge to that disease?

Iframe Embed Code

<iframe src="https://dmccreary.github.io/bioinformatics/sims/biomedical-kg-anatomy/main.html"
        height="560"
        width="100%"
        scrolling="no"></iframe>

Lesson Plan

Grade Level

College introductory bioinformatics

Duration

15-20 minutes

Prerequisites

Understanding of genes, proteins, and their roles in disease
Basic graph theory (nodes, edges, types)
Concept of knowledge representation

Activities

Exploration (5 min): Identify all five node types by color. For each type, find one example and read its details. How many relationships does each node have?
Path Tracing (5 min): Find a drug and trace its path to a disease through a gene target. Write out the full path using the format: Drug → targets → Gene → associated_with → Disease.
Discussion (5 min): Drug repurposing uses knowledge graphs to find existing drugs that might treat new diseases. How could you use this graph to identify a drug repurposing candidate? What kind of path would you look for?
Assessment (5 min): Answer the reflection questions below.

Assessment

What makes a knowledge graph "heterogeneous" compared to a simple network?
Why are labeled edges important in a biomedical knowledge graph? Give an example of two edges that connect the same node types but have different meanings.
How could a bioinformatician use a knowledge graph to predict new gene-disease associations?
What is the difference between the "targets" and "treats" relationships in this graph?