HPEPDOCK

HPEPDOCK performs blind protein-peptide docking to predict binding modes of protein-peptide complexes using a hierarchical algorithm that accounts for peptide flexibility.

Key Features:

• Hierarchical Algorithm: Employs a hierarchical approach to process peptide conformations for docking without requiring extensive refinement simulations.
• Peptide Flexibility Consideration (MODPEP): Incorporates an ensemble of peptide conformations generated by the MODPEP program to capture peptide flexibility during docking.
• Blind Global Peptide Docking Performance: Achieves a 33.3% success rate in predicting binding modes within the top 10 models on a benchmark of 57 unbound cases, compared to pepATTRACT's 21.1% under similar conditions.
• Docking Against Homology Models: Demonstrates a 29.8% success rate within the top 10 predictions when docking against homology models.
• Local Peptide Docking Efficiency: Achieves a 72.6% success rate within the top 10 predictions on a benchmark of 62 unbound cases, versus HADDOCK peptide protocol's 45.2%.
• Computational Efficiency: Requires on average 29.8 minutes per global docking job and 14.2 minutes per local docking job.

Scientific Applications:

• Structural Biology: Predicts protein-peptide interactions to elucidate structural dynamics of protein-peptide complexes.
• Drug Discovery: Provides binding-mode predictions relevant to the identification and optimization of peptide-based therapeutics.
• Peptide Therapeutic Design: Supports rational design of peptide drugs by supplying predicted binding poses and accounting for peptide flexibility.

Methodology:

An ensemble of peptide conformations is generated by MODPEP and processed by a hierarchical docking algorithm to predict global and local binding modes.