PepVis: An integrated peptide virtual screening pipeline for ensemble and flexible docking protocols

Area of Science:

• Computational chemistry
• Drug discovery
• Bioinformatics

Background:

• Peptide therapeutics offer specificity, safety, and cost-effectiveness for treating diseases.
• High-throughput computational screening is crucial for advancing peptide therapeutics in the postgenomic era.
• Existing virtual screening pipelines lack automation and integration for ensemble and flexible docking protocols.

Purpose of the Study:

• To develop PepVis, a GUI-based pipeline for automated peptide modeling and virtual screening.
• To integrate various open-source tools for efficient large-scale peptide screening.
• To address the need for automated and optimized virtual screening pipelines in peptide drug discovery.

Main Methods:

• Developed PepVis, a GUI pipeline integrating Modpep, Gromacs, AutoDock Vina, ZDOCK, AutoDock CrankPep, ZRANK2, and FlexPepDock.
• Utilized Modpep and Gromacs for peptide structure modeling and optimization.
• Employed AutoDock Vina, ZDOCK, and AutoDock CrankPep for virtual screening, with ZRANK2 for rescoring and FlexPepDock for refinement.

Main Results:

• Benchmarking demonstrated that ModPep + Vina outperformed ModPep + ZDock in detecting near-native structures from the LEADS-PEP dataset using ensemble docking.
• PepVis successfully automates large-scale peptide modeling and virtual screening.
• The pipeline is modular and designed for future integration of additional docking algorithms.

Conclusions:

• PepVis provides an efficient, automated solution for peptide-based virtual screening.
• The pipeline enhances the identification of potential peptide therapeutics.
• PepVis is freely available, promoting accessibility in computational drug discovery research.