Inexpensive Method for Selecting Receptor Structures for Virtual Screening.

Area of Science:

• Computational chemistry
• Drug discovery
• Structural biology

Background:

• Virtual screening is crucial for identifying drug candidates.
• Selecting optimal experimental structures for virtual screening is challenging.
• Existing performance metrics can be computationally expensive.

Purpose of the Study:

• Introduce a novel Screening Performance Index (SPI) for selecting effective experimental structures.
• Reduce computational costs associated with virtual screening structure selection.
• Improve the identification of active compounds from virtual screening campaigns.

Main Methods:

• Developed the SPI based on docking energies of known actives to experimental structures.
• Avoided using large decoy sets to minimize computational expense.
• Evaluated SPI performance across eight diverse protein systems (e.g., FABP4, BRAF, ADRB1).

Main Results:

• SPI demonstrated strong agreement with established metrics like BEDROC, RIE, AUAC, and EF.
• SPI proved more effective than docking energy, ligand volume, or crystal resolution in selecting structures.
• The index offers a computationally cheaper alternative for structure selection.

Conclusions:

• SPI is a valuable tool for selecting optimal protein structures for virtual screening.
• It offers a cost-effective and efficient approach compared to traditional methods.
• SPI aids in ensemble docking and situations with limited structural data or time.