AA-Score

AA-Score predicts protein-ligand binding affinities using an amino acid–specific empirical scoring function to support virtual screening, lead optimization, and docking evaluations.

Key Features:

• Empirical Scoring Function: Implements an empirical scoring function tailored for protein-ligand interactions.
• Amino Acid–Specific Interaction Components: Decomposes interactions into amino acid-specific terms including hydrogen bonds, van der Waals forces, and electrostatic interactions.
• Comprehensive Interaction Coverage: Incorporates additional interaction types such as hydrophobic interactions, π-stacking, π-cation interactions, and metal-ligand interactions.
• Performance Evaluation: Evaluated on newly generated test sets that assess scoring, ranking, and docking performance and reported to outperform traditional scoring functions.
• Interaction Fingerprint Analysis and Affinity Prediction: Provides analysis of protein-ligand interaction fingerprints and prediction of binding affinities.

Scientific Applications:

• Virtual Screening: Used to prioritize compounds by predicted binding affinity in early-stage drug discovery.
• Lead Optimization: Applied to refine lead compounds via detailed interaction fingerprints and affinity estimates.
• Docking Studies: Employed to evaluate and rank docking poses and to analyze protein-ligand complexes.

Methodology:

AA-Score applies an empirical scoring approach that decomposes interactions into amino acid-specific components (hydrogen bonds, van der Waals, electrostatic) and includes hydrophobic, π-stacking, π-cation, and metal-ligand terms; performance was assessed using newly generated test sets for scoring, ranking, and docking.