Quokka

Quokka predicts kinase family-specific phosphorylation sites across the human proteome to support analysis of phosphorylation-based post-translational regulation.

Key Features:

• Kinase family-specific prediction: Predicts phosphorylation sites with specificity to kinase families within the human proteome.
• Multiple sequence scoring functions: Integrates multiple sequence scoring functions for characterizing candidate phosphorylation sites.
• Optimized logistic regression: Uses an optimized logistic regression algorithm for classification and scoring of phosphorylation sites.
• Benchmark evaluation: Performance was evaluated using benchmark datasets derived from Phospho.ELM and independent test datasets curated from UniProt.
• Improved predictive performance: Demonstrates significant improvement over current state-of-the-art methods in phosphoproteomic data analysis.

Scientific Applications:

• Phosphorylation site identification: Enables identification of kinase-specific phosphorylation sites for mapping post-translational modifications (PTMs).
• Signaling pathway analysis: Supports analysis of phosphorylation-mediated cellular signaling pathways.
• Protein regulation studies: Facilitates studies of phosphorylation roles in protein degradation and protein-protein interactions.
• Disease-related research: Aids investigation of aberrant phosphorylation linked to human diseases, including cancers.
• Hypothesis generation and validation: Provides predictive data to guide experimental hypothesis generation and biological validation in phosphoproteomics.

Methodology:

Integrates multiple sequence scoring functions with an optimized logistic regression algorithm and was evaluated on Phospho.ELM benchmark datasets and independent UniProt test datasets.