EVEREST: Automatic Identification and Classification of Protein Domains
Author Information
Author(s): Portugaly Elon, Harel Amir, Linial Nathan, Linial Michal
Primary Institution: The Hebrew University of Jerusalem
Hypothesis
Can we automatically identify and classify all protein domains from sequence data?
Conclusion
The EVEREST process provides a comprehensive and validated view of protein domain families, covering 85% of amino acids in the Swiss-Prot database.
Supporting Evidence
- EVEREST defines 20,230 domains, covering 85% of the amino acids of the Swiss-Prot database.
- EVEREST annotates 11,852 proteins that are not annotated by Pfam A.
- Performance tests show that EVEREST recovers 56% of Pfam A families and 63% of SCOP families with high accuracy.
- EVEREST suggests previously unknown domain families with at least 51% fidelity.
Takeaway
This study created a computer program called EVEREST that helps scientists find and group parts of proteins, which are important for understanding how proteins work.
Methodology
The EVEREST process uses machine learning techniques to identify and classify protein domains from sequence data through iterative refinement and statistical modeling.
Potential Biases
Potential biases may arise from the reliance on known domain families for training.
Limitations
The study relies on existing databases and may not capture all possible domain families.
Participant Demographics
The study analyzed protein sequences from the Swiss-Prot section of the UniProt Knowledgebase.
Statistical Information
P-Value
p<0.05
Statistical Significance
p<0.05
Digital Object Identifier (DOI)
Want to read the original?
Access the complete publication on the publisher's website