Fourier and spectral envelope analysis of medically important bacterial and fungal sequences

Abstract

In this paper, we introduce the Fourier and spectral envelope analysis methods to analyze some biomolecular sequences, particularly medically important bacteria and fungi DNA sequences, to get their interesting frequency properties. Fourier analysis includes mapping character strings into numerical sequences, calculating spectra of DNA sequences and setting and solving optimization problem in order to construct a powerful predictor of exons along the long DNA sequences. The spectral envelope analysis makes use of spectral envelope for analyzing periodicities in categorical-valued time series and it is useful for the scaling of non-numeric sequences. The spectral envelope analysis utilizes optimization procedure to improve upon traditional analysis performance in distinguishing coding from non-coding regions in DNA sequences. The two approaches greatly facilitate the understanding of local nature, structure and function of biomolecular sequences. They also provide useful techniques to combine bioinformatics analysis with modern computer power to quickly search for diagnostic patterns within long sequences.