RNSC

RNSC identifies known and predicts novel protein complexes within protein-protein interaction (PPI) networks by partitioning networks using a cost-based clustering algorithm.

Key Features:

• Restricted Neighborhood Search Clustering (RNSC): Explores cluster partitions in PPI networks using a restricted neighborhood search strategy.
• Cost-function optimization: Applies a tailored cost function to optimize the detection of true protein complexes.
• Network partitioning: Partitions protein-protein interaction networks into clusters for downstream analysis.
• MIPS-derived filters: Defines filters based on functional and graph-theoretic properties derived from true protein complexes in the MIPS database to distinguish clusters from complexes.
• Complex identification and prediction: Identifies known protein complexes and predicts potential undiscovered complexes from PPI data.
• Large-scale PPI handling: Operates on large-scale PPI datasets to analyze extensive interaction networks.

Scientific Applications:

• Protein complex identification: Detection and validation of known protein complexes within PPI networks.
• Novel complex prediction: Prediction of potential undiscovered protein complexes from network structure.
• Model organism PPI analysis: Applied to PPI networks from Saccharomyces cerevisiae, Drosophila melanogaster, and Caenorhabditis elegans.

Methodology:

Uses the Restricted Neighborhood Search Clustering algorithm to partition networks via a cost function and applies filters based on functional and graph-theoretic properties derived from true protein complexes listed in the MIPS database.