DynaBiS: A hierarchical sampling algorithm to identify flexible binding sites for large ligands and peptides

Area of Science:

• Computational biology
• Structural bioinformatics
• Drug discovery

Background:

• Identifying protein binding sites is essential for drug discovery but experimentally challenging.
• Computational methods often neglect protein and ligand flexibility, impacting accuracy.
• Mutual structural adaptations significantly influence protein-ligand interactions.

Purpose of the Study:

• To present DynaBiS, a novel hierarchical sampling algorithm for flexible binding site identification.
• To address the limitations of existing methods by incorporating protein and ligand flexibility.
• To improve the accuracy of binding site prediction, especially for large and flexible ligands.

Main Methods:

• DynaBiS employs a hierarchical sampling algorithm inspired by the DynaDock flexible docking algorithm.
• Soft-core potentials are utilized to allow protein-ligand overlap, enabling efficient sampling of conformational changes.
• The algorithm was evaluated against common binding site identification methods using a diverse set of 26 proteins.

Main Results:

• DynaBiS demonstrates superior performance in identifying protein binding sites compared to other evaluated methods.
• The algorithm is particularly effective for large and flexible ligands, including peptides.
• Successful identification of binding sites was achieved using both holo (ligand-bound) and apo (unbound) protein structures.

Conclusions:

• DynaBiS offers a significant advancement in computational binding site identification.
• The explicit consideration of protein and ligand flexibility enhances prediction accuracy.
• This method holds promise for guiding drug discovery efforts, especially for challenging targets.