You are here

Give diferent chain information for a single sequence (comparative modeling)

5 posts / 0 new
Last post
Give diferent chain information for a single sequence (comparative modeling)
#1

I am trying to obtain a comparative model for a fusion protein. The query sequence have two repeats of a single motif that form a monomeric structure. The template structures are, by the way, homodimeric proteins that have this motif in two separate chains. I would like to make the model giving the relative structural information that this two chains have in the homodimer. I do not know how to said this to rosetta in the alignment file:

Query sequence:

------XXXXXXXXXXXXXXXXXXX----------XXXXXXXXXXXXXXXX--------

Template (chain A)

------YYYYYYYYYYYYYYYYYYY----------------------------------

Template (chain B)

----------------- -----------------YYYYYYYYYYYYYYYY--------

The information of the the two chains (homodimer) are in the same PDB file. It would be very helpful if someone could give me a clue of how to write this in the grishin’s alignment format, or other compatible Rosetta alignment format.

Post Situation: 
Tue, 2013-04-02 08:59
Martin Floor

The alignment should look like this:

------XXXXXXXXXXXXXX----------XXXXXXXXXXXXXX--------

Template (chain A)

------YYYYYYYYYYYYYY--------------------------------------------

Template (chain B)
-----------------------------------------YYYYYYYYYYYYYY---------

I hope that you get the idea.

Tue, 2013-04-02 09:04
Martin Floor

The current comparative modeling program in Rosetta only can use one template. So to model based on two templates, you'd have to do two runs. Once for one half, and then a second for the other half. You'd then have to do something like docking to combine the two.

However, the new "hybrid protocol" for comparative modeling will be able to simultaneously model both portions of the protein. I believe (though don't quote me) that it should be released with the upcoming 3.5 release. It is already available on the Robetta server (http://robetta.bakerlab.org/ under "Domain Parsing & 3-D Modeling"). However, Robetta is set up to do things automatically, so there isn't as much control over template choice and alignment tweaking as there would be with a local run.

Tue, 2013-04-02 11:11
rmoretti

I forwarded this along to some of the comparative modeling folks. I don't have a clue.

Tue, 2013-04-02 11:12
smlewis

Thanks for the quick response. I'll be looking forward the new release.
Greetings.

Tue, 2013-04-02 20:37
Martin Floor