Property-based design of GPCR-targeted library.

Area of Science:

• Medicinal Chemistry
• Computational Chemistry
• Drug Discovery

Background:

• G protein-coupled receptors (GPCRs) are crucial drug targets.
• Identifying selective GPCR ligands from large compound libraries is challenging.
• Existing methods for ligand identification can be time-consuming and resource-intensive.

Purpose of the Study:

• To develop a computational method for classifying molecules as potential GPCR ligands.
• To create a scoring scheme to prioritize compounds for GPCR-targeted libraries.
• To accelerate the identification and development of novel GPCR-active drugs.

Main Methods:

• A scoring scheme was designed to classify molecules into "GPCR-ligand-like" and "non-GPCR-ligand-like" categories.
• Molecular structures were encoded using a set of descriptors.
• A neural network was trained to classify molecules based on these descriptors, using extensive GPCR and non-GPCR active agent databases.

Main Results:

• The methodology effectively distinguishes between GPCR and non-GPCR active agents.
• Validation using large datasets confirmed the method's ability to profile molecular requirements.
• The approach enables efficient qualification or disqualification of potential GPCR ligands.

Conclusions:

• The developed scoring scheme and classification method are valuable tools for GPCR-targeted library design.
• This approach aids in the selection and prioritization of potential GPCR ligands for bioscreening.
• The method can significantly constrain library sizes, accelerating the development of new GPCR-active therapeutics.