- Article -
| In Silico Biology 4, 0013 (2004); ©2004, Bioinformation Systems e.V. |
Rapid detection of conserved regions in protein sequences using wavelets
Arun Krishnan*, Kuo-Bin Li and Praveen Issac
Bioinformatics Institute, 30 Biopolis Street, 07-01, Matrix, Singapore 138671
Email: arun@bii.a-star.edu.sg, kuobin@bii.a-star.edu.sg, praveeni@bii.a-star.edu.sg
* corresponding author
Edited by E. Wingender; received November 14, 2003; accepted January 06, 2004; published February 08, 2004
Abstract
We present an algorithm to detect protein sub-structural motifs from primary sequence. The input to the algorithm is a set of aligned multiple protein sequences. It uses wavelet transforms to decompose protein sequences represented numerically by different indices (such as polarity, accessible surface area or electron-ion integration potentials of the amino acids). The numerical representation of a protein sequence has significant correlation with its biological activity, thus common motifs are expected to be observable from the wavelet spectrum. The decomposed signals are then up-sampled and similarity search techniques are used to identify similar regions across all the proteins at multiple scales. Results indicate that wavelet transform techniques are a promising approach for rapid motif detection.
Key words: protein motif detection, wavelet analysis, conserved motifs