Classifying Protein Structures by Their Functional Domains
Author Information
Author(s): Yu Xiaojing, Wang Chuan, Li Yixue
Primary Institution: Bioinformatics Center, Shanghai Institutes for Biological Sciences
Hypothesis
Can the quaternary structure of proteins be classified based on their functional domain composition?
Conclusion
The domain composition approach is a more effective method for predicting protein quaternary structures compared to traditional methods.
Supporting Evidence
- The overall success rate for the domain composition method was 75.17% in the training dataset.
- In an independent dataset, the method achieved an overall accuracy of 84.11%.
- The amino acid composition method achieved only 41.42% accuracy in the same dataset.
Takeaway
This study shows that we can figure out how proteins are built by looking at their parts, which helps us understand their functions better.
Methodology
The study used a nearest neighbor algorithm to classify proteins based on their functional domain composition, validated with jackknife cross-validation.
Limitations
The study is limited to homo-oligomers and may not apply to hetero-oligomers.
Digital Object Identifier (DOI)
Want to read the original?
Access the complete publication on the publisher's website