NetworKIN

NetworKIN predicts in vivo kinase-substrate relationships by integrating consensus motifs with kinase and phosphoprotein network context to assign phosphorylation sites to specific kinases.

Key Features:

• Augmented predictions: Integrates motif-based predictions with kinase and phosphoprotein network context to enhance assignment of in vivo substrate specificity, contributing approximately 60%–80% to predictive power.
• Improved accuracy: Achieves a reported 2.5-fold improvement in constructing accurate phosphorylation networks compared to consensus-motif-only methods.
• Scalable evaluation strategy: Implements a scalable strategy to evaluate and rank the likelihood of predicted kinase-substrate relationships.

Scientific Applications:

• DNA damage signaling: Applied to elucidate kinase-substrate dynamics, including identification of CDK1 as the kinase for 53BP1 and ATM as the kinase for Rad50.
• Substrate prediction: Predicted BCLAF1 as a substrate of GSK-3.

Methodology:

Combines consensus motif–based predictions with network context analysis of kinases and phosphoproteins and uses a scalable evaluation strategy to assess prediction likelihoods.