fPOP: footprinting functional pockets of proteins by comparative spatial patterns

Area of Science:

• Structural bioinformatics
• Computational biology
• Protein science

Background:

• Protein functional surfaces, specifically binding sites, are crucial for molecular interactions.
• Analyzing the geometric shapes of these surfaces can reveal functional patterns.
• Databases like SplitPocket have cataloged functional surfaces from bound protein structures.

Purpose of the Study:

• To develop and present the fPOP database, a comprehensive resource for protein functional surfaces.
• To utilize geometric analysis and spatial templates for predicting binding surfaces in unbound proteins.
• To provide a framework for studying protein conformational changes, functional divergence, and classification.

Main Methods:

• Geometric analysis of protein structures to identify functional surfaces (split pockets).
• Utilizing the Smith-Waterman algorithm for shape comparison and footprinting of unbound pocket fragments against the SplitPocket database.
• Pairwise alignment of unbound and bound pocket fragments to assess local structural similarity via geometric matching.

Main Results:

• The fPOP database now contains approximately 90,000 identified or predicted functional surfaces.
• A method is established for predicting binding surfaces of unbound protein structures using spatial templates from bound forms.
• The approach facilitates the study of protein functional surfaces, conformational dynamics, and functional divergence.

Conclusions:

• The fPOP database is a valuable, accessible resource for researchers studying protein functional surfaces and binding sites.
• The developed methodology enables prediction of binding surfaces in unbound proteins, enhancing our understanding of protein function.
• This work offers a novel framework for classifying proteins based on their functional surface characteristics.