Community Detection on a Pharma Clinical-Trial Investigator Graph
Overview
What this challenge is about.
You receive a pre-fetched dump of around 15,000 trials from a public registry covering oncology over the last 10 years and a mapping of trials to investigator names + institutions. Construct an investigator graph (nodes = investigators, edges = co-authorship on a trial). Apply Louvain and Leiden community detection; pick one based on a modularity-vs-stability comparison. Label the top 30 communities with descriptive summaries (specialty, geography, throughput). Build a visual atlas (interactive Gephi or Streamlit) and write a methodology memo for medical-affairs leadership.
The Brief
What you'll do, and what you'll demonstrate.
Map the global oncology investigator landscape with graph-based community detection to accelerate trial-site selection.
Earning criteria — what you'll demonstrate
- Construct meaningful graphs from public registry data
- Apply Louvain and Leiden community-detection algorithms
- Characterize communities qualitatively and quantitatively
- Visualize large graphs for non-technical stakeholders
Program Fit
Where this fits in your program.
Sharpens the same skills your degree expects you to demonstrate.
Skills
Skills you'll demonstrate.
Each one shows up on your verified credential.
Careers
Roles this prepares you for.
Real titles. Real skill bridges. Pick the one closest to your trajectory.
Data Scientist
Mapping a real-world domain via graph community-detection and shipping a stakeholder-ready atlas is exactly the day-one work of a data scientist at any pharma-AI firm.
This challenge sharpens
- community-detection
- graph-analysis
- network-visualization
Data Engineer
Building reproducible graph-construction pipelines from messy public data is core data-engineering work in knowledge-intensive companies.
This challenge sharpens
- graph-analysis
- python
- community-detection
Applied AI Scientist
Combining algorithmic choices (Louvain vs. Leiden) with operational delivery (atlas + memo) is the applied-AI-scientist craft for analytics-heavy teams.
This challenge sharpens
- louvain
- leiden
- community-detection