ProtFus

ProtFus identifies and characterizes protein-protein interactions associated with fusion proteins in cancer by mining scientific literature and integrating validated interaction data.

Key Features:

• Text Mining Approach: Employs text mining of scientific literature to extract evidence of protein-protein interactions (PPIs) related to fusion proteins.
• Chimeric Protein-Protein Interactions (ChiPPI): Incorporates the ChiPPI method to predict and characterize chimeric PPIs and their effects on cellular networks.
• Validation and Prediction: Uses an online literature search for validation and a Naïve Bayes classifier trained on a set of 358 fusion proteins with corresponding PPIs to identify reliable evidence.
• Comprehensive Screening: Demonstrated capability to screen large datasets, identifying 2,908 PPIs across 18 cancer types from a test set of 1,817 fusion proteins.

Scientific Applications:

• Network Alteration Analysis: Enables study of how fusion proteins alter protein interaction networks in various cancers.
• Genomic–Phenotype Correlation: Supports investigation of relationships among genomic features, disease pathology, and drug metabolism in cancer patients.
• Therapeutic Insight: Provides validated molecular interaction data to inform tailored therapy approaches based on fusion-protein-specific interactions.
• Mechanism Exploration: Aids exploration of mechanisms generating fusion proteins, including chromosomal aberrations and trans-splicing events.

Methodology:

Applies text mining to extract PPI evidence from literature, validates predicted interactions via online literature searches, incorporates the ChiPPI approach, and refines predictions using a Naïve Bayes classifier trained on a 358-fusion-protein dataset.