RoKAI

RoKAI infers kinase activity from phospho-proteomic data by integrating functional networks and propagating phosphosite quantifications to capture coordinated signaling changes.

Key Features:

• Heterogeneous network: Constructs a network with kinases and phosphosites as nodes to enable contextual analysis of phosphorylation events.
• Data integration: Integrates protein-protein interactions, kinase-substrate annotations, co-evolutionary patterns, and structural distances between phosphosites.
• Quantification propagation: Propagates mass spectrometry-derived phosphosite quantifications across the network to capture coordinated signaling changes.
• Input data: Operates on mass spectrometry-derived phospho-proteomic data.
• Robustness and expanded coverage: Enhances prediction accuracy and reliability even with incomplete annotations or quantifications.
• Local phosphorylation signal: Uses the predictive power of phosphorylation events near a kinase to improve activity inference.

Scientific Applications:

• Disease-associated kinase dysregulation: Identifies dysregulated kinases associated with diseases such as cancer, Alzheimer’s disease, and Parkinson’s disease.
• Kinase activity profiling and target prioritization: Provides insights into kinase activities and aids identification of understudied kinases for potential therapeutic intervention.

Methodology:

Constructs a heterogeneous network of kinases and phosphosites, integrates protein-protein interactions, kinase-substrate annotations, co-evolutionary patterns, and structural distances between phosphosites, and propagates mass spectrometry-derived phosphosite quantifications across the network; computational experiments demonstrate that phosphorylation events near a kinase are predictive of its activity.