A graph-based approach to construct target-focused libraries for virtual screening.

Area of Science:

• Computational chemistry
• Drug discovery
• cheminformatics

Background:

• Virtual screening is cost-effective but often yields low hit rates.
• Existing compound libraries are vast and difficult to navigate for novel bioactives.
• Targeted library design is crucial for efficient drug discovery.

Purpose of the Study:

• To develop an automated computational method for synthesizing novel drug-like molecules.
• To create tailored screening libraries for specific protein targets.
• To address the limitations of traditional virtual screening approaches.

Main Methods:

• Developed eSynth, an exhaustive graph-based search algorithm.
• Decomposed organic compounds into rigid fragments and flexible linkers.
• Reconnected molecular fragments based on atomic connectivity patterns.

Main Results:

• eSynth successfully reconstructed over 80% of active molecules from fragments.
• Successfully generated novel scaffolds by reconstructing 40% of target molecules from distinct fragments.
• Demonstrated computational efficiency, generating 90% of molecules within a minute.

Conclusions:

• eSynth reconstructs chemically feasible molecules from fragments.
• Generates diverse molecular collections with desired activity profiles for targeted drug discovery.
• Provides a freely available tool to accelerate the development of novel screening libraries.