Binding site characterization - similarity, promiscuity, and druggability.

Area of Science:

• Computational chemistry and structural biology.
• Drug discovery and medicinal chemistry.

Background:

• Understanding protein binding site similarities is key for polypharmacology, off-target identification, and drug repurposing.
• Promiscuous binding, where ligands interact with multiple targets, can be studied via ligand-based or binding site-based approaches.

Purpose of the Study:

• To apply automated protein binding site comparison methods to a novel dataset.
• To analyze promiscuous binding using sites with shared ligand properties.
• To establish a dataset for benchmarking binding site comparison methods.

Main Methods:

• Automated analysis and comparison of protein binding sites.
• Utilized a novel dataset of protein-ligand complexes with shared shape and physicochemical properties.
• Investigated pocket-centered analyses for promiscuity and druggability.

Main Results:

• Demonstrated the suitability of the novel dataset for benchmarking binding site comparison tools.
• Identified promising avenues for in-depth analyses of binding site promiscuity.
• Highlighted limitations in current binding site similarity assessment and druggability prediction.

Conclusions:

• The developed dataset facilitates the evaluation of new binding site comparison methods.
• Pocket-centered analysis offers valuable insights into drug promiscuity and druggability.
• Awareness of common pitfalls is crucial for accurate binding site analysis in drug discovery.