Biological Sequence Analysis

Our goal is to develop new tools and techniques for biological sequence analysis based on the application of several promising new signal processing technologies. In particular, we hope to develop faster, more accurate, and more robust methods for biological sequence comparison, more compact representation and storage techniques for derived nucleotide or amino acid sequences, and more accurate procedures for the prediction of site functionality and protein molecular structure. In Phase I of this effort, we will restrict our attention to the study of wavelet transform techniques for sequence representation and molecular structure prediction. Wavelet transforms have received a great deal of attention recently and have been shown to be superior to traditional Fourier-based analysis techniques in many respects. Of particular interest for sequence analysis applications are finite-field wavelet transforms. The finite-field wavelet transforms are ideally suited for the analysis of signals (or sequences) that take values in a finite field rather than the fields of real or complex numbers. Since any encoding of a nucleotide or amino acid sequence can be represented quite naturally in a finite field, this technique is very appealing.