Eigenvalue Ratios Reveal Shared Binding Pocket Shapes in RNA and Protein Structures.

Area of Science:

• Structural Biology
• Computational Chemistry
• Drug Discovery

Background:

• Molecular recognition is crucial for drug design, requiring precise characterization of ligand-binding pockets.
• Protein binding sites are well-studied, but the geometric properties of RNA pockets are less understood.
• Existing methods often lack a unified approach for comparing RNA and protein binding site geometries.

Purpose of the Study:

• To develop a unified, size-independent geometric framework for describing and comparing RNA and protein binding pocket shapes.
• To enable direct comparison of pocket geometries across different macromolecule types without specific assumptions.
• To classify binding pockets into archetypal shapes for a more organized analysis.

Main Methods:

• Developed size-independent and residue-agnostic geometric measures of global anisotropy.
• Applied the framework to analyze 300 RNA and 300 protein binding pockets.
• Defined four pocket shape archetypes: sphere-like, rod-like, disk-like, and strongly anisotropic.

Main Results:

• A largely shared geometric landscape exists between RNA and protein binding pockets.
• Sphere-like pockets are more common in proteins, while disk-like and strongly anisotropic pockets are enriched in RNA.
• Rod-like pockets show comparable frequencies in both RNA and proteins; strongly anisotropic pockets are significant in both.

Conclusions:

• The developed framework provides a transferable geometric reference for comparative analysis of RNA and protein pocket architectures.
• This approach reduces structural heterogeneity by organizing diverse binding sites into reproducible geometric regimes.
• The findings support enhanced exploration of RNA pocket accessibility in structure-based drug design studies.