#DOTD: profile Hidden Markov model

19 Jun 2017

A profile Hidden Markov model (profile-HMM) is an acyclic linear HMM that contains three types of states: match, insert and delete, corresponding to symbol frequencies, insertions and deletions, respectively.

 

The information needed to build profile-HMMs is from Multiple Sequence Alignments (MSAs) of protein families. The use of profile-HMMs, therefore, is in modelling protein families to allow for efficient querying of unannotated protein sequences.

 

The below diagram shows how profile-HMMs are built from MSAs.

 

 This blog post is based on the following paper:

 

Yoon, B.J., 2009. Hidden Markov models and their applications in biological sequence analysis. Current genomics, 10(6), pp.402-415.

 

[169]

 

Share on Facebook
Share on Twitter
Share on LinkedIn
Please reload

Please reload

Related Posts
PhDomics by Fatima