Statistical models called hidden Markov models are a recurring theme in computational biology. What are hidden Markov models, and why are they so useful for so many different problems?