Network2Canvas

Network2Canvas visualizes large biological networks on a square toroidal canvas to reveal local connectivity patterns and enriched gene- or drug-set signatures.

Key Features:

• Toroidal Canvas Visualization: Places network nodes on a square toroidal canvas to preserve adjacency relations and represent complex network topology.
• Scalability: Applicable to networks exceeding approximately 100 nodes and 200 links to improve visibility of local connections.
• Simulated Annealing for Node Clustering: Uses simulated annealing to optimize node placement by maximizing local connections and clustering related nodes.
• Node Brightness and Local Fitness: Encodes each node's brightness proportional to its local fitness, reflecting the density of nearby connections.
• Enrichment Highlighting and Clustering Metrics: Highlights enriched terms on the canvas and computes the degree of clustering for input lists of genes or drugs.

Scientific Applications:

• Human and mouse gene-set visualization: Visualizes networks derived from human and mouse gene-set libraries and highlights enriched terms within those networks.
• FDA-approved drug-set visualization: Visualizes networks derived from FDA-approved drug-set libraries and highlights enriched drug-set terms.
• Gene-set and drug-set enrichment analysis: Computes and analyzes clustering degree for lists of genes or drugs to detect biological signatures.
• Systems biology investigations: Supports exploration of genetic interactions and identification of potential drug targets through network-pattern analysis.

Methodology:

Network data are mapped onto a square toroidal canvas, a simulated annealing algorithm clusters nodes by connectivity to optimize layout, node brightness is computed from local fitness (connection density), and enriched terms are highlighted with computation of clustering degree.