Introduction to RISK¶
RISK (Regional Inference of Significant Kinships) is a next-generation tool for biological network annotation and visualization. It integrates advanced clustering, statistically rigorous overrepresentation analysis, and high-resolution plotting to uncover structured relationships in complex networks.
RISK is designed to:
- Identify biologically meaningful modules in large-scale networks
- Perform fast and flexible overrepresentation testing
- Generate publication-ready visualizations
- Scale to hundreds of thousands of edges on standard hardware
- Generalize beyond biology to interdisciplinary networks
What makes RISK different?¶
| Feature | RISK |
|---|---|
| Clustering | Louvain, Leiden, Markov, Spinglass, etc. |
| Statistics | Permutation, hypergeometric, chi-squared, z, etc. |
| File Formats | NetworkX, Cytoscape (.cys, .cyjs), GPickle |
| Annotation | JSON, CSV, Excel, TSV, Python dict |
| Scalability | Networks with 500k+ edges on standard hardware |
| Visual Output | High-resolution SVG, PNG, PDF, etc. |
| Beyond Biology | Used on physics citation network (Supp. Fig. S9) |
Example Applications¶
- Functional module identification in yeast PPI and GI networks
- Synthetic lethality pathway mapping
- Drug-target prediction via annotation overrepresentation
- Cross-domain network exploration in physics and other disciplines
For more examples, see 7. Advanced Plotting.
Full Interactive Example¶
For a full interactive workflow, see tutorial.html, which walks through all key steps—from loading a network to generating publication-quality figures—in a single notebook. You can also download the notebook and data as a ZIP.
The notebook focuses on a curated yeast PPI network (Michaelis et al., 2023), filtered to retain proteins with six or more interactions (2,059 nodes, 27,690 edges). Throughout the tutorial, we showcase:
- Network clustering and module detection
- Annotation-based overrepresentation analysis
- Interactive and advanced visualization features
- Exporting parameters and reproducibility tools
This notebook complements the Markdown documentation in this repository, but does not follow a strict 1-to-1 mapping. Use it for end-to-end examples and contextual guidance.
You can also launch the Quickstart notebook on Binder for a fully interactive experience without installing anything locally.
Getting Started¶
Continue to 1. Installation to install RISK and set up your environment.