Substructure-based virtual screening for adenosine A2A receptor ligands.

Area of Science:

• Medicinal Chemistry
• Computational Chemistry
• Pharmacology

Background:

• Adenosine receptor (AR) ligands are crucial for various therapeutic applications.
• Existing drug discovery methods include target-based and ligand-based virtual screening.
• Comparing different virtual screening strategies is essential for optimizing ligand discovery.

Purpose of the Study:

• To design and evaluate a virtual ligand-based screening approach for novel A(2A) adenosine receptor (AR) ligands.
• To compare the efficacy of ligand-based virtual screening against a previously published target-based approach.

Main Methods:

• Frequent substructure mining was used to derive distinguishing structural features from known A(2A) AR antagonists.
• Multiple screening models were built, with the best statistical model selected for large-scale screening.
• A commercial vendor library was screened using the optimized ligand-based model.

Main Results:

• 36 candidate compounds were selected for experimental testing.
• Eight compounds (22% hit rate) significantly inhibited radioligand binding at A(2A) AR.
• The ligand-based approach identified new ligands distinct from those found by target-based screening.

Conclusions:

• The developed virtual ligand-based screening approach is effective for discovering novel A(2A) AR ligands.
• This method provides a valuable alternative or complementary strategy to target-based screening.
• The study highlights the potential of substructure mining for building robust virtual screening models.