Pepper

Pepper identifies protein complexes by expanding seed protein lists into densely connected subnetworks within proteome-wide protein–protein interaction networks using multi-objective optimization of coverage and density.

Key Features:

• Seed-based expansion: Begins from user-provided seed lists of proteins, typically derived from proteomic studies, to generate candidate complexes.
• Multi-objective optimization: Simultaneously optimizes two objectives—coverage and density—to select candidate complexes.
• Coverage objective: Maximizes inclusion of proteins from the initial seed list within each identified complex.
• Density objective: Enforces high interconnectivity among proteins in a solution based solely on interactions present in a comprehensive proteome-wide interaction network.
• Post-processing pipeline: Provides an automated post-processing stage that supports topological analysis and integration of additional data related to predicted proteins.
• Benchmarking with gold standards: Has been evaluated via comparative analyses using gold standard datasets from yeast and human, reporting superior performance over traditional methods.

Scientific Applications:

• Protein complex identification: Detects densely connected protein subnetworks representing candidate protein complexes from proteomic seed lists.
• Proteome-scale interaction analysis: Explores proteome-wide protein–protein interaction networks to assess interconnectivity and complex plausibility.
• Method benchmarking: Enables comparative evaluation against other complex-detection methods using yeast and human gold standard datasets.
• Integration of additional protein data: Facilitates incorporation of supplementary protein-related data into complex interpretation via post-processing.

Methodology:

Starts from seed protein lists and applies multi-objective optimization to maximize coverage (seed inclusion) and density (edge interconnectivity) using a proteome-wide protein–protein interaction network, followed by automated post-processing for topological analysis and integration of additional protein data.