Pocket-space maps to identify novel binding-site conformations in proteins

Area of Science:

• Computational chemistry
• Structural biology
• Medicinal chemistry

Background:

• Novel binding-site conformations are crucial for structure-based ligand design.
• Exploring diverse pocket shapes enhances drug discovery by addressing potency, selectivity, toxicity, and pharmacokinetics.

Purpose of the Study:

• To introduce PocketAnalyzer(PCA), an automated method for diverse pocket selection.
• To demonstrate the utility of PocketAnalyzer(PCA) in identifying novel binding-site conformations and rationalizing protein-ligand interactions.

Main Methods:

• PocketAnalyzer(PCA) applies principal component analysis and clustering to grid-based pocket detection output.
• The method directly uses pocket shape descriptors, avoiding issues with proxy descriptors.
• Simultaneous analysis of mutants, isoforms, and multi-source protein structures is supported.

Main Results:

• PocketAnalyzer(PCA) identified novel, computationally derived binding-site conformations for aldose reductase and viral neuraminidase.
• These novel conformations were not previously observed crystallographically.
• Known inhibitors were identified that exploit these novel binding-site conformations.

Conclusions:

• PocketAnalyzer(PCA) is a valuable tool for structure-based drug design.
• The approach enhances understanding of the molecular basis of protein-ligand interactions and bioactivity.
• The PocketAnalyzer(PCA) software is available as an open-source Python program.