Hi Careina,
In answer to your first question you could also try the iPATCH server:
http://portal.stats.ox.ac.uk/userdata/proteins/i-Patch/home.pl
This takes two reference structures for proteins that interact, and combined
with multiple sequence alignments of their homologs attempts to predict the
surface contact residues between them.
As far as your second question is concerned, a quick google search using the
term "protein interaction prediction from sequence" gave some useful links, one
of which is Struct2Net:
http://groups.csail.mit.edu/cb/struct2net/webserver/
This tool attempts to predict protein-protein interactions purely from sequence
data. However, it does use a structure-based threading approach, so your
sequences will be run against the pdb. If they are unique to anything in the
structural databases, it may not be useful.
Hope this helps,
Mohammad
----------------------------------------
Dr. Mohammad W. Bahar
Division of Structural Biology
Wellcome Trust Centre for Human Genetics
University of Oxford
--- Begin Message ---
Hi Careina,
For the first question, it sounds as though IBIS would do what you want:
http://www.ncbi.nlm.nih.gov/Structure/ibis/ibis.cgi. It will also try to
answer your second question, although sequences are compared to known
structures, so if your sequence is dissimilar to anything in the PDB it
won't work. It looks as though you can only put in one query
sequence/structure and will then have to scan the results for the
appearance of the second.
I hope this helps.
-- David
On 6 September 2012 07:43, Careina Edgooms <careinaedgo...@yahoo.com> wrote:
> Anybody know of any software out there that can predict potential
> interaction sites between two proteins? They have been shown to interact
> via y2h screens but I have no idea if they would interact on their own in
> vitro. Before I clone them into a vector and purify them I would like some
> sort of confirmation that the interaction could occur in the absence of
> other cellular factors.
> There are 2 interactions I am looking at. For the one, the structures of
> both proteins are known. For the other only one structure is known.
> So, is there software that uses 2 known structures to predict binding
> sites and (I know this is a long shot), but is there any software around
> that could predict an interaction based on the sequences only (or one 3D
> structure and one sequence)?
>
> Thanks
> Careina
>
--- End Message ---
