[Statistical characteristics of primary structures of the functional regions of the Escherichia coli genome. III. Computer recognition of coding regions]

Abstract

We have presented the method for recognition of structural domains of DNA. This method uses statistical description of coding and non-coding regions in the form of stationary or nonstationary Marcov chain, which was introduced in our previous papers. Calculation of the probability that the given fragment of the DNA appears part of the coding region, is the main operation of this algorithm. The results, obtained for the number of E. coli DNA sequences showed the ability of the method to find the structural domains and correct reading frame, so as to give the estimation of the extent of protein expressivity. Provided necessary statistical data are available, the proposed method may be used for the analysis of DNA of other organisms.