Related Experiment Videos
Protein domain identification and improved sequence similarity searching using PSI-BLAST
Richard A George1, Jaap Heringa
1Division of Mathematical Biology, National Institute for Medical Research, The Ridgeway, Mill Hill, United Kingdom.
Proteins
|September 5, 2002
Summary
DOMAINATION accurately predicts protein structural domain boundaries. This method improves the detection of distant homologies in multidomain proteins compared to standard PSI-BLAST searches.
Area of Science:
- Bioinformatics
- Computational Biology
- Structural Biology
Background:
- Protein homology searches are challenged by multidomain proteins, which can lead to premature search termination or excessive results.
- Identifying individual structural domains within a protein sequence is crucial for accurate functional and evolutionary analysis.
Purpose of the Study:
- To develop and evaluate DOMAINATION, a novel method for inferring protein structural domain boundaries.
- To improve the sensitivity and accuracy of homology searches for multidomain proteins.
Main Methods:
- DOMAINATION infers protein domains and boundaries using local gapped alignments from PSI-BLAST.
- It incorporates a novel technique to identify domain insertions and permutations.
- Delineated domains are queried iteratively in subsequent database searches.
Main Results:
- DOMAINATION predicts structural domain boundaries with 50% accuracy across 452 multidomain proteins.
- The method enhances the discovery of distant homologies by 14% compared to standard PSI-BLAST.
- A web-based tool for DOMAINATION is publicly available.
Conclusions:
- DOMAINATION effectively addresses the challenges posed by multidomain proteins in homology searches.
- The tool improves the identification of evolutionary relationships by accurately delineating protein domains.
- This advancement offers a valuable resource for researchers in bioinformatics and structural biology.