Correlation Network

[bann...@gmail.com]

Hi all,

I just recently started using cytoscape and I'm running into some trouble building a correlation network.  My data includes pearson correlation values for a single source gene to a list of target genes and is in the following format:

Source Gene       Target genes       Pearson Correlation

geneX                  geneA                 0.8

geneX                  geneB                 0.7

geneX                  geneC                 0.9

and I'm trying to create a correlation network where edge length would be inversely proportional to the correlation values (so more highly correlated genes would be visualized closer together and less correlated genes would be farther apart). Any advice on how I could do this in a visually appealing way?

Thank you so much!

Manal

[alex.pico]

Hi Manal,

If you haven't already done this, the first step is to do File>Import>Network from file and specify you source, target and edge attribute columns. Then you can choose the "edge weighted" layouts, specifying your "Pearson Correlation" column as having the weight values.  This will give you a layout that takes weights into consideration. Note that an *exact* 2D solution for edge weights expressed as length is often not possible once your network includes cycles, etc.

 - Alex