AlphaSpace2

AlphaSpace2 analyzes biomolecular surface topography to identify and characterize concave regions such as small-molecule binding pockets and protein–protein interaction interfaces.

Key Features:

• β-Cluster Pocket Representation: Uses β-cluster representations as pseudomolecular models of fragment-centric pockets that capture pocket shape and atomic characteristics.
• Pocket–Ligand Shape Comparison: Enables direct comparison between biomolecular concavities and potential ligand structures.
• β-Score Ligandability Metric: Calculates a β-score defined as the optimal Vina score of a β-cluster to estimate binding pocket ligandability.
• Ensemble β-Cluster Mapping: Applies an ensemble β-cluster approach to enable one-to-one mapping and comparison of pockets across aligned protein structures.

Scientific Applications:

• Binding Site Identification: Detects and characterizes small-molecule binding pockets in biomolecular structures.
• Protein–Protein Interaction Analysis: Examines concave interface regions involved in protein complex formation.
• Fragment-Based Ligand Design: Supports optimization of fragment-based ligands through structural comparison of pockets and ligand shapes.

Methodology:

AlphaSpace2 models biomolecular concavities using β-cluster representations, evaluates ligandability with β-scores derived from optimal Vina scores, and performs ensemble β-cluster mapping to compare pockets across aligned protein structures.