CARNIVAL

CARNIVAL derives causal network architectures from gene expression footprints by integrating prior knowledge sources and solving network identification as an integer linear programming problem.

Key Features:

• Prior Knowledge Integration: Integrates signed and directed protein-protein interactions, transcription factor targets, and pathway signatures.
• TF Weight Assignment: Assigns transcription factor weights from PROGENy scores.
• Optimization Flexibility: Optimization parameters can be freely adjusted.
• Multiple Solution Aggregation: Multiple solutions can be obtained and aggregated.
• Efficient Computation: Implementation as an integer linear programming (ILP) problem guarantees efficient computation.

Scientific Applications:

• Signaling Network Contextualization: Contextualizes large signaling networks from gene expression data to identify upstream cellular processes and regulators.
• Disease Mechanism Investigation: Generates hypotheses on potential upstream alterations that propagate through signaling networks, as demonstrated in IgA nephropathy where specific signaling pathways including Wnt and TGF-β were identified and experimentally validated.

Methodology:

CARNIVAL formulates network identification as an integer linear programming (ILP) problem and integrates prior knowledge including signed and directed protein-protein interactions, transcription factor targets, and pathway signatures to derive causal network architectures from gene expression footprints.