Motifs are overrepresented sequence or spatial patterns appearing in proteins. They often play important roles in maintaining protein stability and in facilitating protein function. When motifs are located in short sequence fragments, as in transmembrane domains that are only 6-20 residues in length, and when there is only very limited data, it is difficult to identify motifs. In this study, we introduce combinatorial models based on permutation for assessing statistically significant sequence and spatial patterns in short sequences. We show that our method can uncover previously unknown sequence and spatial motifs in \beta-barrel membrane proteins and that our method outperforms existing methods in detecting statistically significant motifs in this data set. Last, we discuss implications of motif analysis for problems involving short sequences in other families of proteins.
|Number of pages||13|
|Journal||IEEE/ACM Transactions on Computational Biology and Bioinformatics|
|State||Published - 2010|
- Combinatorial algorithms
- Permutations and combinations
- combinatorial models
- sequence analysis.
- short sequence