PCSF

PCSF performs network-based integration of heterogeneous high-throughput biological data to identify subnetworks of interacting proteins and genes relevant to experimental signals.

Key Features:

• Implementation: Provided as an R package implementing the Prize-collecting Steiner Forest (PCSF) approach.
• Algorithm: Uses a Prize-collecting Steiner Forest graph optimization approach to select subnetworks from input networks.
• Data mapping: Maps heterogeneous high-throughput data onto biological networks, including protein-protein interactions, gene-gene interactions, and correlation or coexpression-based networks.
• Subnetwork identification: Identifies high-confidence subnetworks using interaction networks as templates and predicts functional units within the biological system.
• Functional enrichment analysis: Performs functional enrichment analysis on identified subnetworks to assess biological significance.

Scientific Applications:

• Complex disease mechanism elucidation: Integrates high-throughput data with networks to provide insights into mechanisms underlying complex diseases.
• Therapeutic target identification: Enables identification of potential therapeutic targets by highlighting subnetwork components associated with experimental signals.
• Functional module discovery: Supports discovery of functionally coherent modules within protein and gene interaction networks.

Methodology:

Maps high-throughput data onto interaction networks and applies the Prize-collecting Steiner Forest graph optimization to identify high-confidence subnetworks, followed by functional enrichment analysis.