Automated Protein Subfamily Identification and Classification
Author Information
Author(s): Brown Duncan P, Krishnamurthy Nandini, Sjölander Kimmen
Primary Institution: University of California, Berkeley
Hypothesis
Can a computational pipeline improve the accuracy of protein subfamily identification and classification?
Conclusion
The SCI-PHY algorithm significantly enhances the classification of proteins into subfamilies, improving specificity and reducing errors in functional annotation.
Supporting Evidence
- SCI-PHY subfamilies correspond closely to functional subtypes defined by experts.
- Subfamily HMMs improve the separation between homologous and non-homologous proteins.
- Extensive validation shows high specificity in classification of novel sequences.
Takeaway
This study created a computer program that helps scientists figure out what different proteins do by grouping them into families based on their similarities.
Methodology
The study used a computational pipeline that includes de novo subfamily identification and hidden Markov models for classification.
Limitations
The methods may be sensitive to alignment errors and the definitions of subfamilies can be somewhat arbitrary.
Digital Object Identifier (DOI)
Want to read the original?
Access the complete publication on the publisher's website