SPServer

SPServer evaluates protein structures and protein-protein interactions using split-statistical (knowledge-based) potentials to provide residue-level and interface-level scoring for structural assessment.

Key Features:

• Decomposition of Statistical Potentials: Split-statistical potentials are decomposed into distinct terms based on amino acid pair types, hydrophobicity, solvent accessibility, and secondary structure type.
• Comprehensive Scoring Functions: The method computes global scores and residue-residue pair profiles and presents residue-level scores as score plots and maps.
• Identification and Analysis Capabilities: The approach identifies potentially problematic regions within protein structures, pinpoints disrupting amino acid pairs at protein interfaces, and enables comparison of tertiary and quaternary structural models.
• Unique Integration: Fold (tertiary) and protein-protein interaction (quaternary) assessments are integrated within a single scoring framework.
• Local Quality Assessment: Provides residue-level assessment of structure and interface quality to support comparative analyses across models.

Scientific Applications:

• Protein Structure Prediction: Identifies near-native structures among predicted models using statistical potential scoring.
• Protein Docking: Ranks protein docking poses based on interface and residue-level scores.
• PPI Binding Affinity Prediction: Supports prediction of protein-protein interaction binding affinities using knowledge-based potential-derived scores.

Methodology:

SPServer applies split-statistical (knowledge-based) potentials derived from empirical data, decomposes them by amino acid pair types, hydrophobicity, solvent accessibility, and secondary structure, and computes global scores plus residue-residue score profiles presented as plots and maps to assess protein folds and interfaces at the residue level.