automating the determination of a core genome set from annotated genomes

[alanmcn1]

Dear all

I am currently working on some large data sets where I have assembled and annotated closely related bacterial genomes. In order to identify "core" and "accesory" genes within my strain set I have been csowly working through artemis and tagging genes as orthologues. This becomes extremely time consuming as the number of genomes increases. Does anyone know of an automated process for this? I have considered using Mauve or Mugsy but to my knowledge these just allow extraction of blocks of sequence, not actual orthologous genes. In essence what i want is a txt file list of genes present in all the strains, and genes unique to strains.

Any help would be greatly appreciated

Alan

[joscarhuguet]

Orthomcl does a good job. The software predicts orthology based on reciprocal blast. The output gives you a general picture of what you want. However for more specific details you want to reconstruct some trees with the putative orthologs. The software does not take in account gene synteny, because it needs only the coding sequences. In my hands most of the predicted orthologues show a clear synteny though.
Here is the link:

http://www.orthomcl.org/cgi-bin/OrthoMclWeb.cgi

[nickloman]

Alan, if you supply Genbank files as Mauve input so it knows where the genes are you will find it offers the ability to export an orthologue report. It's a menu option.

[alanmcn1]

Many thanks Gents. I will check out both options tomorrow. It's problems like this that make me think i need to find a really good Perl course