Building networks using Biogrid plugin and a list of genes
82 views
Skip to first unread message
Louise Laurent
Jan 28, 2016, 9:26:32 PM1/28/16
to cytoscape-helpdesk
Hi,
I would like to import a list of genes, and build a network using the Biogrid plugin. I have installed the Biogrid plugin, such that "BIOGRID-ORGANISM-Homo_sapiens-3.4.129.mitab" is displayed under the network tab. I have generated a txt file with the gene names (in official genesymbols) in column A and "1" for each gene in column B, as suggested in the response to a previous post; this file was imported using File-->import-->table-->to selected networks only-->node table columns-->shared name-->case sensitive key names/unchecked. A new column in the "Table Panel" appeared with the heading "1". However, the rest of the column is empty. Is it possible to use the imported gene list to select nodes from the Biogrid network to build a network? In addition, is it possible include "back nodes" (nodes that are not on my list, but that are "between" nodes on my list -- e.g. which would enable the construction of connections like this: [node from my list]-[edge from Biogrid]-[node from Biogrid]-[edge from Biogrid]-[node from my list])?
Scooter Morris
Feb 4, 2016, 10:19:20 AM2/4/16
to cytoscape-helpdesk
Hi Louise,
Sounds like you're using Cytoscape 2.8. I think the first step would be to try to use Cytoscape 3.3, which is our current release. BioGrid networks are available directly from the Welcome Screen. Once you've read in the network, note that the "name" column contains the internal BioGrid (Entrez Gene) identifier. You will need to map that to some useful identifier (this is probably the problem you're having in the older version, also). One easy way to do that is to install the "bridgeDB" app, which will allow you to map from the BioGrid identifier to any of a number of useful identifiers. Alternatively, you could create your input file to use the BioGrid identifier rather than gene symbol. Once you have selected the desired set of proteins, the next step would be to look at Pesca, which calculates the shortest path between nodes. That might allow you to create the network you want.
An alternatively approach might be to load your network from STRING (using the StringApp) rather than BioGrid. Still doesn't solve the intermediate node problem, but it may be a shortcut to getting the initial network.
-- scooter
Sounds like you're using Cytoscape 2.8. I think the first step would be to try to use Cytoscape 3.3, which is our current release. BioGrid networks are available directly from the Welcome Screen. Once you've read in the network, note that the "name" column contains the internal BioGrid (Entrez Gene) identifier. You will need to map that to some useful identifier (this is probably the problem you're having in the older version, also). One easy way to do that is to install the "bridgeDB" app, which will allow you to map from the BioGrid identifier to any of a number of useful identifiers. Alternatively, you could create your input file to use the BioGrid identifier rather than gene symbol. Once you have selected the desired set of proteins, the next step would be to look at Pesca, which calculates the shortest path between nodes. That might allow you to create the network you want.
An alternatively approach might be to load your network from STRING (using the StringApp) rather than BioGrid. Still doesn't solve the intermediate node problem, but it may be a shortcut to getting the initial network.
-- scooter
William Redwine
Feb 19, 2016, 2:22:24 PM2/19/16
to cytoscape-helpdesk
Louise,
Did you find an easy way to build your network? I think I have a similar problem, which is to take data from a proteomics screen I performed and map it onto preexisting networks (such as BioGrid). I am looking to see how connected the hits from my screen are. I am really new to Cytoscape, so if you found something that works well, please share.
-Bret
Reply all
Reply to author
Forward
0 new messages