Comparative genomics via wavelet analysis for closely related bacteria (Q1773680)

Summary: Comparative genomics has been a valuable method for extracting and extrapolating genome information among closely related bacteria. The efficiency of the traditional methods is extremely influenced by the software method used. To overcome this problem, we propose using wavelet analysis to perform comparative genomics. First, global comparisons using wavelet analysis gives the differences at a quantitative level. Then local comparisons using keto-excess or purine-excess plots show precise positions of inversions, translocations, and horizontally transferred DNA fragments. We firstly found that the level of energy spectra difference is related to the similarity of bacteria strains; it could be a quantitative index to describe the similarities of genomes. The strategy is described in detail by comparisons of closely related strains: S.typhi CT18, S.typhi Ty2, S.typhimurium LT2, H.pylori 26695, and H.pylori J99.