Three dimensional shape comparison of flexible proteins using the local-diameter descriptor

Area of Science:

• Structural Bioinformatics
• Computational Biology
• Biophysics

Background:

• Protein function inference relies heavily on shape similarity comparisons.
• Existing methods often treat proteins as rigid bodies, failing to account for flexibility.
• Protein flexibility is crucial for biological processes, necessitating methods that capture shape dynamics.

Purpose of the Study:

• To introduce an efficient approach for comparing flexible protein shapes.
• To adapt the Local Diameter (LD) descriptor for capturing invariant properties of protein backbone motion.
• To evaluate the LD descriptor's effectiveness in protein shape retrieval compared to conventional methods.

Main Methods:

• Adapted the Local Diameter (LD) descriptor to analyze 3D protein shapes.
• Assigned a local diameter value to each surface point, capturing neighborhood shape.
• Constructed a 1D histogram from diameter distributions for shape representation and comparison.

Main Results:

• The LD descriptor accurately accounts for protein shape deformations.
• Protein shape comparison was simplified to a distance calculation between 1D feature vectors.
• LD descriptor demonstrated superior performance in consistency and discernibility over conventional descriptors.

Conclusions:

• The LD descriptor offers an effective technique for comparing flexible protein shapes.
• The method is insensitive to protein shape deformation, proving robust.
• Potential applications include molecule retrieval by shape and rapid protein structure retrieval.