[Bioperl-l] SeqHound

[Brian Osborne]

Raul,

Good question about BIND, I don't know if the public version is up-to- 
date. For the latest protein-protein interaction data I look at public  
databases like IntAct.

If you use bioperl-network you should be able to read IntAct data into  
a graph, then find the interactions that you're interested in in that  
graph. There are a few qualifications to this statement though, like  
do you have the "right" identifiers or names. So you're right, bioperl- 
network constructs graphs from the XML files but the interactions you  
want are in those graphs. Something like:

  my $graphio = Bio::Network::IO->new(-file   => 'human.xml',
                                      -format => 'psi25');
  my $graph = $graphio->next_network();
  my $node = $graph->get_nodes_by_id('UniProt:P12345');
  my @neighbors = $graph->neighbors($node);



Brian O.

On Sep 11, 2008, at 7:32 AM, Raul Mendez Giraldez wrote:

> Hi Brian,
>
> Actually I realized later that SeqHound is a part of Bioperl itself,  
> and
> that regarding BIND (at least the public database) is reachable trough
> BIND SOAP protocol, that can be implemented in perl through the module
> SOAP::Lite. I still don't know whether the public BIND database is out
> of date, or which part of the BOND database it covers.
>
> Regarding the Bio::Network packages, at the Bioperl suite, I guess it
> rather for representing protein - protein interaction graphs, isn't  
> it?
> That could be interesting to me, but in a second step. I am more
> concerned now in getting this protein - protein interaction data,  
> for a
> set of proteins some biologists gave me. I don't know anything about
> Cytoscape, normally I'm trying to exploit perl data management
> capabilities.
>
> Thanks for the info.
>
> Cheers,
>
> Raul
>