Virtual Screening for HIV Protease Inhibitors Using a Novel Database Filtering Procedure.

Area of Science:

• Computational chemistry
• Drug discovery
• Structural biology

Background:

• Virtual screening is crucial for identifying potential drug candidates.
• HIV protease inhibitors are vital for antiretroviral therapy.
• Efficient database filtering is essential for large-scale virtual screening.

Purpose of the Study:

• To develop and validate a novel feature matrix matching procedure for virtual screening.
• To identify novel inhibitors for HIV protease from a large chemical database.
• To reduce the computational cost of drug discovery through effective ligand preselection.

Main Methods:

• A structure-based approach was used to analyze the HIV protease active site, creating a graph representation.
• A novel feature matrix matching procedure was applied to filter a large ligand database (ZINC).
• Ligands were preselected based on graph comparison and subsequently subjected to molecular docking.

Main Results:

• The filtering procedure reduced approximately 14 million ligands to a subset of 14,299.
• The method successfully identified both novel and previously known experimentally validated ligands.
• The approach demonstrated high efficiency in preselecting potentially active compounds.

Conclusions:

• The novel feature matrix matching procedure is a valid and efficient method for virtual screening.
• This approach can significantly accelerate the identification of drug candidates, including HIV protease inhibitors.
• The study highlights the utility of graph-based methods in computational drug discovery.