Augustus: Choosing Species

Question

[emmajosenl]

Hi All

I'm working with Agustus to predict genes. The question is what is the criteria to identify the closest related species among the list of already predicted genes. http://bioinf.uni-greifswald.de/augustus/ says "the human version is good for all mammals" since mammals are closely related species. Let me explain this with my scenario: I am trying to predict genes using Agustus on bacterium "Bradyrhizobium ganzhouense". There are 3 bacteria species already predicted according to the above site.

1. Escherichia coli

2. Staphylococcus aureus

3. Thermoanaerobacter tengcongensis

Which is the best species that can be used with Augustus from the above list to predict "Bradyrhizobium ganzhouense"? What is the criteria for choosing that? How can I identify the closest species among the above 3 to bacterium "Bradyrhizobium ganzhouense" which can be used with Augustus?

Any help appreciated and thanks in advance.

[GenoMax]

AUGUSTUS is a program that predicts genes in eukaryotic genomic sequences.

Is there a particular reason you want to use AUGUSTUS?

Since you are working with a bacterium shouldn't gene prediction be much simpler (other possible tools) prodigal, GeneMark?

Answer 1

[piet]

Which is the best species that can be used with Augustus from the above list to predict "Bradyrhizobium ganzhouense"?

you can compare higher order taxonomic ranks like Family, Order, Class. Genes associated with basic bacterial metabolism like RNA processing, protein expression, DNA duplication, cell division, energy conversion can be expected to be conserved in phylogenetically related species. But this assumption must not be true for genes involved in specialized secondary metabolism, which have often been aquired by horizontal gene transfer and are likely to have no orthologs in phylogenetically related species.

Answer 2

[emmajosenl]

Thank you so much for the reply.

But for eukaryotes can I take an organism from one class which can be used as trained species for all others coming under the same Class. http://bioinf.uni-greifswald.de/augustus/ says "Note that for closely related species usually only one version is necessary" For example there are 10 trained species for Class Insecta. Can I take one version from the list below which can be used for all?

1. Acyrthosiphon pisum
2. Aedes aegypti
3. Apis mellifera
4. Bombus terrestris
5. Culex pipiens
6. Drosophila melagonaster
7. Heliconius melpomene
8. Nasonia vitripennis
9. Rhodnius prolixus
   1. Tribolium castaneum

[GenoMax]

Please use ADD REPLY/ADD COMMENT when responding to existing posts. That helps to keep threads logically organized.