Combining Protein Sequence and Structure Alignments
Author Information
Author(s): Gundolf Schenk, Thomas Margraf, Andrew E. Torda
Primary Institution: Centre for Bioinformatics, University of Hamburg
Hypothesis
Can a single framework effectively align protein sequences and structures?
Conclusion
The proposed probabilistic framework effectively merges sequence and structure descriptors for protein alignment.
Supporting Evidence
- The framework allows for fast structural alignments.
- It can identify cases where sequence and structural similarities disagree.
- The method is tolerant of poor structures.
- It can handle unusual protein structures effectively.
Takeaway
This study shows a new way to compare proteins by looking at both their sequences and structures together, which can help find similarities that might be missed if only one is used.
Methodology
The study used a probabilistic approach to classify protein fragments based on their sequence and structural properties.
Potential Biases
The underlying statistical models may not perfectly represent protein data.
Limitations
The method may not find the optimal classification or the correct number of classes for realistic problems.
Digital Object Identifier (DOI)
Want to read the original?
Access the complete publication on the publisher's website