Searching ECOD for Homologous Domains by Sequence and Structure

Area of Science:

• Bioinformatics
• Structural Biology
• Evolutionary Biology

Background:

• The Protein Data Bank (PDB) contains numerous protein structures, but organizing their evolutionary domains is complex.
• Existing structural domain resources may not be continuously updated or prioritize homology for classification.

Purpose of the Study:

• To describe the Evolutionary Classification Of Domains (ECOD) database.
• To highlight ECOD's unique approach to classifying protein evolutionary domains based on homology.
• To explain how to access and utilize ECOD for bioinformatics research.

Main Methods:

• ECOD employs a hybrid approach, using automated protein-protein BLAST for most classifications.
• Manual curation, including alignment, functional analysis, and literature review, is used for ambiguous cases.
• The database is continuously updated, classifying thousands of proteins weekly.

Main Results:

• ECOD provides a continuously updated classification of evolutionary domains from PDB structures.
• The database prioritizes homology over sequence or structural similarity for domain organization.
• ECOD facilitates functional inference by placing proteins within their evolutionary context.

Conclusions:

• ECOD offers a valuable, up-to-date resource for understanding protein evolution and function.
• Its homology-based classification scheme enhances the interpretation of protein structures.
• The database is accessible via the web and provides data files for large-scale use.