Orthomcl-results without co-orthologs and in-paralogs

Question

[natasha.sernova]

Dear all,

Could you, please, help me? I was running Orthomcl-program, using a small bacterial taxon. I used their *.faa versions for each genome, just proteins.

I was going to find bona fide orthologs only, but occasionally I realized that as a result I have a mixture of bona fide orthologs, co-orthologs and in-paralogs! Is GETHOGs my only way to find bona fide orthologs, or there are some other procedures to get rid of these co-orthologs and in-paralogs in the Orthomcl output? I thought, the "named groups" from Orthomcl is a solution I was looking for, but unfortunately I was wrong.

Thank you very much!

Sincerely yours,
Natasha

Answer 1

[Christophe Dessimoz]

Indeed, OrthoMCL builds groups of orthologs and "close paralogs". If you need groups containing orthologs only, you could use the OMA standalone version (http://omabrowser.org/standalone) on your data and look at the resulting OrthologousGroups files.

GETHOGs is not the right solution for your problem. It infers hierarchical orthologous groups (HOGs), which are groups of genes that descended from a common ancestral gene in a taxonomic range of interest. It however does not matter whether members of a HOG are related through speciation (=orthologs) or duplication (=paralogs), so HOGs may contain either.

[natasha.sernova]

Dear Dr.Dessimoz,

Many thanks for your answer!

May I contact you by e-mail - there are too many details to discuss,

and I couldn't find any active e-mails in your papers?

Mine is this one: [e-mail]

Thank you very much!

Sincerely yours,
Natasha