Entering edit mode
7.8 years ago
mirza
▴
180
Hi,
I need to create a large network for my organism of interest as it is not there in cytoscape. So, I downloaded network data (.xls and .csv files) from public network databases for my organism, imported the files in cytoscape, created individual networks and then went for "merge networks." Now, I am stuck there for more than 24hrs now (The memory button is green!). I am using Cytoscape 3.4, in Dell Tower workstation with a 256 GB RAM. My cytoscape.vmoptions file shows -Xms2048M and -Xmx248690M.
Cross-posted on Cytoscape-discuss list. Did you try the suggestions on the discussion i.e. remove duplicate edges and self loops ?
@Jean-Karim Hi, Yes, I tried this, for both edges and self loops but I am not sure if I understood it correctly. This has to be done before merging then I did it by selecting the networks both individually and simultaneously. Both times it showed 0 duplicate edges and 0 self loops If it has to be done after merger, then....I am stuck at the merger step for 3 days now.
No one is responding in Cytoscape-discuss and helpdesk :(
It's better to create your own question than to tag along someone else's. Also I think the question might be more appropriate for the helpdesk. See the Cytoscape community page.
How big are your networks ? How many of them do you have ? Also make sure that they use the same identifiers for the nodes. If Cytoscape doesn't work for you, you could try the union function in igraph or build the union graph with the perl Graph module.
I am trying to create a rice network so that I can analyze my genes using it. I have network data sheets of 14Mb (.csv) and 562 Mb (txt) from the public databases. 328 Mb (rice.dbxref). I get the problem, they are all different forms and formats of data. Any suggestions??
If you can import them into Cytoscape, the problem is likely not the file formats. What you should check is that all the genes/nodes are referred to using the same identifiers in all the networks, e.g. make sure that one network doesn't refer to nodes using Uniprot accessions while another is using rice gene names of the form LOC_Osxxx.
ok, thanks a lot. If dats the case and if I convert and make the identifiers common, will it work? or it will complicate the job?
And I have a few basic questions, these might sound lame to you but I am really new to networking: I have transcriptome data, I have up-regulated and down-regulated genes and have calculated the fold changes. 1. Can I go ahead and make interaction network with this data or do I need to do some other calculations or analysis, before importing it in Cytoscape 2. I have my data in .xls and .csv format, do I need to convert it in some other format? and 3. Cytoscape asks to define "source" and "target" set of genes (columns), how should I arrange my data then?
Most likely. If you don't do this, Cytoscape has no way of knowing which genes/nodes are common across all networks.
You need your data in a format supported by Cytoscape. The simplest format is SIF. How you convert your up/down-regulated genes into a network is up to you and depends on what you want the links to represent.
Not if it is already in a supported format.
See above.
thanks a lot Jean, I hope I can bother you again :)
Hi, I did go through this manual earlier. What I need help with is how to actually arange my data, make an .xls or .csv file which will work in these tool, I mean how to I don't have protein protein interaction data as shown in the example the manual. Do you know any other detailed tutorials or workshop links for cytoscape or in general for networking?