Application of hidden Markov models for the identification of short protein repeats

In this paper, hidden Markov models (HMMs) are discussed in the context of molecular biological sequence analysis. The statistics relevant in the HMM approach are described in detail. An HMM based method is used to analyze two proteins that contain short protein repeats (SPRs). As a benchmark, a state-ofthe- art program for the detection of SPRs is also used for both proteins. Finally, an outlook for combination possibilities of HMMs with phylogenetic approaches is given.