Exploring the Structural Lexicon of the Proteome via Metric Geometry

Area of Science:

• Structural biology
• Computational biology
• Bioinformatics

Background:

• Protein structure dictates function, but systematic comparison across the proteome is challenging.
• Existing methods lack comprehensive frameworks for organizing and comparing diverse protein structures.

Purpose of the Study:

• Introduce GWProt, a novel computational framework for protein structure alignment and analysis.
• Leverage metric geometry, specifically Gromov-Wasserstein couplings, for enhanced structural comparisons.
• Integrate biochemical information and quantify local conformational differences.

Main Methods:

• Developed GWProt, a computational framework utilizing metric geometry principles.
• Applied Gromov-Wasserstein couplings for protein structure alignment.
• Introduced a measure for local geometric distortion to capture conformational variations.

Main Results:

• Identified conformational switches within proteins.
• Detected conserved functional domains in distantly related viral proteins.
• Revealed topological rearrangements in homologous protein folds.
• Uncovered recurrent structural motifs in human proteome functional domains.

Conclusions:

• Metric geometry offers a versatile and quantitative approach for systematic protein structure analysis.
• GWProt complements existing methods for protein organization and functional studies.
• The framework facilitates deeper understanding of protein structure-function relationships.