Finding pathogenicity islands and gene transfer events in genome data

Motivation: There is a growing literature on wavelet theory and wavelet met hods showing improvements on more classical techniques, especially in the c ontexts of smoothing and extraction of fundamental components of signals. G +C patterns occur at different lengths (scales) and, for this reason, G+C p lots are usually difficult to interpret. Current methods for genome analysi s choose a window size and compute a chi (2) statistics of the average valu e for each window with respect to the whole genome. Results: Firstly, wavelets are used to smooth G+C profiles to locate charac teristic patterns in genome sequences. The method we use is based on perfor ming a chi (2) statistics on the wavelet coefficients of a profile; thus we do not need to choose a fixed window size, in that the smoothing occurs at a set of different scales. Secondly, a wavelet scalogram is used as a meas ure for sequence profile comparison; this tool is very general and carl be applied to other sequence profiles commonly used in genome analysis. We sho w applications to the analysis of Deinococcus radiodurans chromosome I, of two strains of Helicobacter pylori (26 695, J99) and two of Neisseria menin gitidis (serogroup B strain MC58 and serogroup A strain Z2491). We report a list of loci that have different G+C content with respect to the nearby re gions; the analysis of N. meningitidis serogroup B shows two new large regi ons with low G+C content that are putative pathogenicity islands.