SurvNet

SurvNet identifies network-based biomarkers by integrating biological networks, molecular profiling (e.g., microarray expression data, DNA methylation data), and patient survival data (e.g., progression-free survival) to find subnetworks whose activity correlates with patient survival.

Key Features:

• Input integration: Integrates three input files: a biological network file, a molecular profiling file, and a patient survival data file.
• Network types supported: Accepts biological network representations such as gene regulatory networks and protein interaction networks.
• Molecular data types: Processes high-throughput molecular profiling data including microarray expression data and DNA methylation data.
• Survival metrics: Uses patient survival data metrics such as progression-free survival for association analysis.
• Parameterization: Employs user-defined parameters to guide the search for relevant subnetworks.
• Subnetwork search: Systematically searches for subnetworks within the provided biological network that show strong correlation with patient survival.
• Biomarker output: Produces a list of network biomarkers ranked by their association with patient survival outcomes.
• Network-centric approach: Leverages biological networks as the platform to enhance robustness of biomarker discovery.

Scientific Applications:

• Prognostic biomarker discovery: Identification of subnetworks predictive of patient survival using gene expression or DNA methylation data.
• Molecular mechanism analysis: Elucidation of survival-associated molecular interactions within gene regulatory and protein interaction networks.
• Clinical stratification: Stratification of patients based on network biomarkers correlated with progression-free survival.
• Translational research: Informing diagnostic and therapeutic research by highlighting survival-associated network modules.

Methodology:

Integrates a biological network file, a molecular profiling file, and a patient survival data file; uses user-defined parameters to systematically search subnetworks within the biological network and identifies those subnetworks whose activity shows the highest correlation with observed patient survival, producing a list of network biomarkers.