Snooker: a structure-based pharmacophore generation tool applied to class A GPCRs.

Area of Science:

• Biochemistry
• Computational Biology
• Pharmacology

Background:

• G-protein coupled receptors (GPCRs) are crucial drug targets for numerous diseases.
• Small molecules modulate GPCR function by binding to the extracellular side of transmembrane domains.
• Developing effective drugs requires understanding these binding interactions.

Purpose of the Study:

• To present Snooker, a structure-based computational method for generating pharmacophore hypotheses.
• To enable drug discovery for GPCRs, particularly for apo-proteins (without known ligands).
• To apply the method across diverse GPCR subfamilies.

Main Methods:

• Construction of homology models for GPCR transmembrane (TM) domains.
• Prioritization of residues based on predicted ligand-binding probability.
• Conversion of protein properties to ligand space for pharmacophore feature generation.
• Application to 15 diverse GPCR targets.

Main Results:

• Snooker pharmacophore hypotheses accurately predicted binding modes for ~75% of compounds for the beta-2-adrenergic receptor.
• Generated hypotheses identified essential residues for ligand binding across 15 GPCRs.
• Virtual screening using Snooker hypotheses yielded high target specificity and significant enrichment factors (>10-fold) for 8 out of 15 targets.
• Prospective predictions for the dopamine D3 receptor were highly ranked in a community-wide docking challenge.

Conclusions:

• Snooker is a versatile, knowledge-driven bioinformatics tool for generating GPCR pharmacophore hypotheses.
• The method is effective for apo-proteins and applicable to the entire GPCR family.
• Snooker facilitates the identification of specific drug candidates and aids in drug discovery for GPCR targets.