CyNetSVM

CyNetSVM integrates gene expression data with protein-protein interaction (PPI) networks using network-constrained support vector machines (NetSVM) to identify network biomarkers and predict clinical outcomes in cancer research.

Key Features:

• Integration with Cytoscape: Operates as a Cytoscape application to connect analyses with biological network visualization and attribute data.
• Gene expression and PPI integration: Integrates gene expression profiles with protein-protein interaction (PPI) networks to enable network-based biomarker discovery.
• Network-constrained Support Vector Machines (NetSVM): Implements NetSVM that constrains support vector machines by incorporating network topology to prioritize biologically coherent features.
• Computational efficiency: Optimizes performance for analyses on large genomic datasets.
• Network visualization capabilities: Visualizes identified subnetworks within biological networks to contextualize biomarkers in signaling pathways.

Scientific Applications:

• Cancer outcome prediction: Identifies subnetworks predictive of clinical outcomes, including breast cancer recurrence, from gene expression and PPI data.
• Network biomarker discovery: Pinpoints biomarkers enriched in signaling pathways relevant to cancer progression to inform prognosis and treatment response.

Methodology:

Integrates gene expression profiles with PPI data to construct a network-based model and applies the NetSVM algorithm, which constrains SVMs by network topology, to identify subnetworks predictive of clinical outcomes and genes central within those networks.