ClusPro

ClusPro performs automated protein–protein docking to predict three-dimensional structures of protein complexes from input coordinate files or PDB IDs retrieved from the Protein Data Bank (PDB).

Key Features:

• Input handling: Accepts coordinate files and PDB IDs and retrieves structures from the Protein Data Bank (PDB).
• Automated docking with FFT: Evaluates billions of candidate complexes using a Fast Fourier Transform (FFT) correlation approach combined with pairwise interaction potentials to identify favorable surface complementarities.
• Energy evaluation and filtering: Filters docking conformations using electrostatic and desolvation free energy criteria to retain energetically favorable models.
• Clustering and ranking: Clusters filtered conformations to identify broad energy wells and ranks clusters according to clustering properties to prioritize putative complexes.
• Refinement techniques: Incorporates Monte Carlo simulations and medium-range optimization methods such as SDU (Simulated Diffusion Umbrella sampling) for initial refinement.
• Benchmark performance: Demonstrated competitive results in CAPRI rounds 13–19 and frequently produces near-native complexes with reported average RMSD around 5 Å for cases with limited backbone conformational change.
• Structure-based pairwise potentials (DARS): Uses Decoys As the Reference State (DARS) and related structure-based intermolecular potentials derived from complex structures and decoy conformations, enhancing performance for enzyme–inhibitor complexes and indicating areas for optimization in antibody–antigen docking.

Scientific Applications:

• Functional genomics: Predicts protein–protein interactions to support mapping of interaction networks and hypothesis generation in functional genomics studies.
• Structural biology: Models complexes from separately determined protein structures to aid interpretation of molecular mechanisms and structural hypotheses.
• Drug discovery and interaction studies: Supports modeling of enzyme–inhibitor complexes and evaluation of candidate interfaces relevant to therapeutic design and lead optimization.

Methodology:

ClusPro applies FFT-based docking with pairwise structure-based potentials (including DARS), filters by electrostatic and desolvation free energies, performs clustering and ranking of conformations, and uses Monte Carlo simulations and SDU for initial refinement.