SHOW stands for Structured HOmogeneities Watcher. It is a set of programs implementing different uses of Hidden Markov Models (HMMs) for DNA sequences. SHOW enables self-learning of HMM on a set of sequences, sequence segmentation based on the Baum-Welch or the Viterbi algorithms, and sequence simulation under a given HMM. These programs are designed to allow the user to specify any highly structured model and also to process large sets of sequences.

The three following programs are available :
  show_emfit enables to fit an HMM on sequences using EM algorithm (learning) and to reconstruct the hidden state path using a forward-backward algorithm (segmentation). When use with fixed parameters, show_emfit only produces the sequence segmentation with the forward-backward algorithm.
  show_viterbi implements the Viterbi algorithm to find the most probable hidden path given the observed sequence (segmentation). The HMM parameters can first be learned with show_emfit.
  show_simul enables to simulate a hidden state sequence and a DNA sequence under a specified HMM.