Design, docking, and evaluation of multiple libraries against multiple targets.

Area of Science:

• Computational chemistry
• Drug discovery
• Structural biology

Background:

• Designing and screening combinatorial libraries is crucial for identifying novel drug candidates.
• Efficiently exploring vast chemical spaces against protein targets remains a challenge.

Purpose of the Study:

• To present a general computational approach for the design, docking, and virtual screening of multiple combinatorial libraries against protein families.
• To validate the method's ability to predict binding modes and experimental data.

Main Methods:

• A three-stage method involving scaffold docking, side-chain substituent selection using a 'divide-and-conquer' algorithm, and library comparison.
• Application to three serine proteases (trypsin, chymotrypsin, elastase) and three combinatorial libraries.
• Utilized a vector-based orientation filter for scaffold docking and a free-energy-based scoring procedure.

Main Results:

• The scaffold docking procedure accurately reproduced crystallographic binding modes.
• The scoring procedure successfully predicted experimental binding data for protease inhibitor mutants.
• The method effectively discriminated between different types of virtual libraries (peptide, benzodiazepine, tetrahydroisoquinolinone).

Conclusions:

• The developed computational method provides an efficient and accurate approach for designing and screening combinatorial libraries.
• The findings have significant implications for optimizing library design in drug discovery.
• The 'divide-and-conquer' algorithm enables linear-time exploration of large substituent lists.