SCOWLP classification: structural comparison and analysis of protein binding regions

Area of Science:

• Structural biology
• Bioinformatics
• Computational biology

Background:

• Understanding protein interactions is crucial for deciphering molecular recognition mechanisms.
• Experimental structures reveal proteins binding diverse ligands, necessitating tools to classify protein binding regions (PBRs).
• A robust PBR classification aids in analyzing protein interaction networks, functions, and rational design.

Purpose of the Study:

• To develop and implement a novel classification system for protein binding regions.
• To enhance the SCOWLP database with detailed PBR classifications.
• To facilitate comparative analysis of PBRs at atomic resolution.

Main Methods:

• Extracted interaction data from the SCOWLP database, including protein domains, peptides, and interfacial solvent.
• Defined a similarity index based on overlapping interacting residues from pairwise structural alignments.
• Employed agglomerative hierarchical clustering with a complete-linkage method for PBR classification at various similarity cut-offs.

Main Results:

• Implemented a hierarchical classification of PBRs within the SCOWLP database, extending the SCOP classification.
• SCOWLP now contains 9,334 binding regions across 2,561 families, with 65% of families exhibiting multiple PBRs.
• 22% of identified PBRs form complexes with multiple protein families, highlighting interaction complexity.

Conclusions:

• The SCOWLP classification provides a framework for studying protein interfaces and comparative PBR analysis.
• Enables detailed examination of interactome conservation and variability at the atomic level.
• The classification is valuable for reconstructing protein complexes, understanding networks, and designing ligands, with the web application available at http://www.scowlp.org.