Related Experiment Videos
Toward consistent assignment of structural domains in proteins.
Stella Veretnik1, Philip E Bourne, Nickolai N Alexandrov
1San Diego Supercomputer Center, University of California, San Diego, 9500 Gilman Dr., La Jolla, CA 92093-0537, USA. veretnik@sd.sc.edu
Journal of Molecular Biology
|May 19, 2004
Summary
Quality assurance for protein domain assignment is lacking. This study evaluates six methods, revealing key factors causing discrepancies and recommending caution for structural domain analysis.
Area of Science:
- Structural biology
- Bioinformatics
- Computational biology
Background:
- Protein domain assignment from 3D structures is crucial for understanding protein evolution and function.
- Limited quality assurance exists for current protein domain assignment methods.
Purpose of the Study:
- To evaluate and compare the performance of six common protein domain assignment methods.
- To identify factors contributing to discrepancies in domain assignments.
Main Methods:
- Comparison of three expert methods (AUTHORS, CATH, SCOP) and three automated methods (DALI, DomainParser, PDP).
- Analysis of individual method performance against author annotations.
- Consensus analysis across groups of methods (expert, automatic, combined).
Main Results:
- Significant differences observed in structural domain assignments across evaluated methods.
- Key factors causing conflicts include small domain definitions, split secondary structures, and complex domain architectures.
- Expert and automated methods exhibit varying levels of agreement and disagreement.
Conclusions:
- Caution is advised when utilizing current protein domain assignment tools.
- Further research and refinement are needed to improve the accuracy and consistency of domain assignment.
- A web-based resource is available for benchmarking and analysis of domain assignments.