This is a test version of Biostars. For the public version, visit https://www.biostars.org.
How to combine Uniprot protein names and STRING interaction table for Cytoscape?

Hello everybody!

I'm new in bioinformatics and I'm starting new project. I have a STRING protein interaction network of some bacteria and Uniprot proteins list of it. I load STRING table in Cytoscape to get the graph representation, but all proteins have names like 272947.RP001 etc. As far as I know that's because STRING uses KEGG names. So how can I add normal uniprot names and short description of molecular function to each protein? Or may be I can get normal names from GeneBank or some other Database?

Thank you for all possible help.

string cytoscape ppi uniprot
• 6,552 views • link
updated by Ram Ram • written by swarekwood • swarekwood

2 answers

Create an attribute file that has mapping between the two ID types, basically a two column table. Upload this table into cytoscape using node names/IDs as keys. This will create another column in the node table with the new ID as an extra node attribute. Then in visual styles, use a passthrough mapper to label the nodes with the new IDs.

• 0 views • link
written by Jean-Karim Heriche Jean-Karim Heriche

Thank you very much for your answer!

Unfortunately, I didn't manage to solve the problem, following your instruction.

I created the table with KEGG protein names in the first column and normal IDs in the second one, While uploading, I'm choosing KEGG names as the key col. Nevertheless, Cytoscape creates another column in the node table, but it's absolutely empty! I tried string and integer values as the ID and none of them were recognized succesfully, What can be a reason of that?

P.S. Sorry for possible mistakes, because English is not my native language and I'm still learning it.

N.

• 0 views • link
written by swarekwood • swarekwood

First make sure you're using cytoscape 3 (I can't remember how things were done in 2). If the column is empty, it's likely an import problem like importing an empty column because of delimiter mismatch. The way to import is File>Import>Table>File. Then check your text file import options (click the 'Show text file import options' checkbox), in particular that you have the delimiter set to the one used in your file.

• 0 views • link
written by Jean-Karim Heriche Jean-Karim Heriche

And thank You very much!

I've found the problem: in uploading table code of specie and protein ID were delimited by ; and by . in Cytoscape. Now that's fully OK and I can start working on my project.

Another one question. if you let. What plugin would you recoomend to regroup nodes to subnetworks based on involved pathway and does this opportunity ever exist in Cytoscape?

• 0 views • link
written by swarekwood • swarekwood

Use the same procedure to add another column with pathway information. Then if you want the nodes to be together in the layout, use either Layout>Attribute circle layout>your pathway column or Layout>Group attributes layout>your pathway column. If you want to use another layout type, then you need to play with edge weights, typically ensuring that edge weights within pathways are higher then edges across pathways. If you want to distinguish pathway nodes by color, use the node style and apply a discrete mapping of colors to pathways.

• 0 views • link
written by Jean-Karim Heriche Jean-Karim Heriche

We are currently developing a proper Cytoscape App for the STRING database. It also makes this easier, since the node attributes table will already have UniProt accession number for the proteins if you fetch net network this way. Also, it preserves the STRING "glass ball" look within Cytoscape.

Although still being developed, you can already find v0.9 here: http://apps.cytoscape.org/apps/stringapp

• 1 views • link
updated by Ram Ram • written by Lars Juhl Jensen Lars Juhl Jensen

Log in to answer this question.

More posts like this