GeneMark Universal gene prediction program – predicts protein-coding regions, in genomes of any type.
MkMat Model building tool for GeneMark. Makes matrices from sample protein-coding and non-coding sequences selected by the user.
Prokaryotic GeneMark.hmm Predicts genes in prokaryotic genomes. Leaves less room for expert than GeneMark
GeneMarkS Self-training version of GeneMark.hmm for prokaryotic genomes.
Eukaryotic GeneMark.hmm Gene prediction program – predicts protein-coding regions in eukaryotic genomes. (exon-intron structures)
Custom models We can build models for a new eukaryotic genome, you name it.

Remark: The difference between GeneMark and GeneMark.hmm program can be described in terms of the theory of Hidden Markov models (HMM)
- the GeneMark algorithm can be approximated by the Forward-Backward algorithm computing local a posteriori probability of protein coding function.
- the GeneMark.hmm program implements Viterbi algorithm for semi Markov HMM to determine a global parse of a DNA sequence into protein coding and non-coding regions.