Things to do with the clustered groups of proteins from orthoMCL tool???

Question

[JstRoRR]

Hi,

I am trying to create a substitution matrix for some bacterial species. To achieve this, as a preliminary step, I have performed a search for orthologs (protein sequences) using orthoMCL. The program outputs a group.txt file containing groups of sequence clustered on the basis of similarity. Now the next step would be to pair wise align orthologs and create a substitution matrix.

My question is, should I consider taking orthologs grouped by the program in the groups.txt file or should I consider paired orthologs present in pairs folder(orthologs.txt) in the output????

[Torst]

Be warned that OrthoMCL does not output the singleton clusters ie. some genes are unique.

Other tools you might want to consider are ProteinOrtho5 and kClust and cd-hit which can do similar things (and much faster).

[JstRoRR]

Thanx Torst for providing alternate tools. I will go through them.

I have one more question, if you have any idea about, is there any tool available which can be used to create a substitution matrix from multiple alignments? Or I have to do it manually or write a fresh script for that?

[amanjain]

OrthoMCL does provides a perl script to extract Singletons

[Naren]

OrthoMCL just gives the list of sequences not present in the groups file. but the accuracy of being a true singleton is skeptical, as these remaining sequences may have some orthologs within them.