Improving Protein Fold Classification with Loop Topology
Author Information
Author(s): Jeong Jieun, Berman Piotr, Przytycka Teresa
Primary Institution: The Pennsylvania State University
Hypothesis
Can the inclusion of loop topology information improve the accuracy of protein fold classification based on secondary structure?
Conclusion
Including information about loop topology significantly improves the accuracy of protein fold classification methods.
Supporting Evidence
- Log-odds scores showed significantly better clustering in 27 cases.
- Improvements were observed even with loop lengths as small as four.
- Different folds responded variably to the inclusion of loop information.
Takeaway
This study found that adding details about loops in proteins helps scientists better understand how proteins are shaped and classified.
Methodology
The study used a set of beta proteins to test the impact of loop topology on protein fold classification, employing local alignment scores and a genetic algorithm for optimization.
Potential Biases
Potential overfitting due to the optimization of parameters based on a limited training set.
Limitations
The improvements in accuracy were not uniform across all protein folds, with some folds experiencing worse performance.
Participant Demographics
The study focused on beta proteins with at most 30% pairwise identity.
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