AlphaSpace 2.0: Representing Concave Biomolecular Surfaces Using β-Clusters.

Area of Science:

• Computational chemistry and structural biology
• Drug discovery and molecular modeling

Background:

• Rational drug design often focuses on concave biomolecular surfaces like binding sites and protein-protein interaction interfaces.
• Accurate topographical mapping of these concavities is crucial for understanding molecular interactions and guiding modulator design.

Purpose of the Study:

• To introduce a novel pseudomolecular representation, the β-cluster, for fragment-centric pockets.
• To enable direct shape comparison between biomolecular pockets and potential ligands for optimized drug design.
• To develop a quantitative measure (β-score) for pocket ligandability and a method for comparing pockets across protein structures.

Main Methods:

• Utilized AlphaSpace, a computational tool for topographical mapping of biomolecular concavities, to detect fragment-centric pockets.
• Developed the β-cluster representation to mimic the shape and atomic details of potential molecular binders.
• Defined the β-score based on optimal Vina docking scores within the β-cluster and implemented an ensemble β-cluster approach for structural comparisons.

Main Results:

• The β-cluster representation facilitates direct pocket-to-ligand shape comparison, aiding in ligand optimization.
• The β-score effectively indicates the ligandability of a given pocket.
• The ensemble β-cluster approach enables accurate one-to-one pocket mapping and comparison across aligned protein structures.

Conclusions:

• The β-cluster is a valuable tool for characterizing biomolecular pockets and guiding fragment-based ligand optimization.
• AlphaSpace 2.0, incorporating β-cluster functionalities, provides a comprehensive platform for binding site analysis and modulator design.
• This approach demonstrates utility in binding site detection, protein-protein interaction characterization, and structure-function studies.