Translate A Nucleotide Alignment To Protein Alignment

Question

[Whetting]

Hi,
I have somewhat of a bizarre question. I want to align a bunch of proteins. However, while they are homologous, the similarity is really poor, making it hard to get a decent alignment. However, my protein of interest is nested within another protein. The larger protein is a two domain protein. Both domains align well, however, the linker connecting these domains does not . My protein of interest is is located on the n+1 frame (so if the larger protein is on frame +1, the protein of interest is on frame +2). Obviously, my protein of interest consists mainly of nucleotides coding for this linker region in the larger protein. I was thinking that I may be able to align the larger protein, reverse translate this alignment to nucleotides and translate the resulting nucleotide alignment into the +2 frame. At this point I should have a better alignment of my protein of interest.
However, does anyone know of a program that will translate a nucleotide alignment (i.e. consider looking for codons across gaps?)

I hope this was clear? please let me know if you need more info!

As always, thanks for the help!

Answer 1

[Dan]

JalView can do this for you. The overlapping reading frames should be highly conserved.

[Whetting]

As far as I know, JalView will only allow for translation in the +1 frame...

Answer 2

[cdsouthan]

This is only a partial answer but I would do TBLASN with the full cDNA of your protein of interest. You will then see protein aligments in all (6) frames and this will gap-cross if the scores on either side of the frame shift are decent. If, what I assume is a real protein linker domain, has some homology/similarity scores you should pick it up, but remember it could be deep in the hit list.