UNIT 4.6 Eukaryotic Gene Prediction Using GeneMark.hmm

  1. Mark Borodovsky1,
  2. Alex Lomsadze2,
  3. Nikolai Ivanov2,
  4. Ryan Mills2

Published Online: 1 MAY 2003

DOI: 10.1002/0471250953.bi0406s01

Current Protocols in Bioinformatics

Current Protocols in Bioinformatics

How to Cite

Borodovsky, M., Lomsadze, A., Ivanov, N. and Mills, R. 2003. Eukaryotic Gene Prediction Using GeneMark.hmm. Current Protocols in Bioinformatics. 1:4.6:4.6.1–4.6.12.

Author Information

  1. 1

    School of Biology and School of Biomedical Engineering, Georgia Institute of Technology, Atlanta, Georgia

  2. 2

    School of Biology, Georgia Institute of Technology, Atlanta, Georgia

Publication History

  1. Published Online: 1 MAY 2003
  2. Published Print: MAR 2003

This is not the most recent version of the article. View current version (1 SEP 2011)


In this unit, eukaryotic GeneMark.hmm is presented as a method for detecting genes in eukaryotic DNA sequences. The eukaryotic GeneMark.hmm uses Markov models of protein coding and noncoding sequences, as well as positional nucleotide frequency matrices for prediction of the translational start, translational termination and splice sites. All these models along with length distributions of exons, introns and intergenic regions are integrated into one Hidden Markov model. The unit describes running the program over the Internet and locally on a Unix machine. It also discusses GeneMarkS EV, which can be used to detect genes in eukaryotic viruses.