Hidden Markov Model (HMM)

Probabilistic model used to describe a succession of states by associating hidden states with observed ones; a Markov Model is used to describe transitions between these hidden states. Such hidden states can be ‘intron’, ‘intergenic’ or ‘exon’ for models predicting gene structure, ‘slow’ or ‘fast’ for models predicting evolutionary rate, or different tree topologies for models predicting gene trees or recombination. (Bousseau 2009) A statistical approach that is used to estimate a series of hidden states (for example, ancestry at loci along a chromosome). The method is based on observations of the states that have uncertainty (for example, the ancestral assignment of sequence reads) and the expected probability of transitions between states (for example, recombination breakpoints). (Davey 2011). statistical models based on Bayesian methods, which use probabilities rather than scores found in profiles. Used for clustering methods based on sequence alignment (Martinez 2011)