Analysis of the structure of genes using hidden Markov models

Background To perform genome analysis, semi-Markov models were widely developed and are efficient to detect protein genes. The trade off is a strong increase of the complexity of most algorithms implied by the estimation and the use of these models. Hidden Markov models (HMMs) are effective tools to detect series of statistically homogeneous structures, but they are not well suited to analyse complex structures. Numerous methodological difficulties are encountered when using HMMs to segregate genes from transposons or retroviruses, or to determine the isochore classes of genes. The aim of this paper is to analyse these methodological difficulties, and to suggest new tools for the exploration of genome data. Results We show that hidden Markov models can be used to analyse complex genes structures with bell-shaped distributed lengths, modelling them by sums of geometric laws. Thus, macros-states model the distributions of the lengths of the regions. Variou