Hello, its Arno Germond again,
I am using the following line of script to draw the network of genes:
net.result <- network(PLSMODEL, cex.node.name = 1, comp = 1:2, cutoff = 0.7, col.names = TRUE, row.names = axis, lty.edge = "solid", lwd.edge = 1, show.edge.labels = FALSE)
I noticed that in the net.result$gR object, the names of the variables (X and Y) were not taken into account, (there are replaced by a number), although that when drawing the network (on plot), the name appeared properly on the network figure (both X and Y when I use the above line of code).
The consequence of this "unlabelled" gR object, is that, when I export the network model into Cytoscape 3 for example, it will result in a network without labels. I could not find a way of "relabelling" the network with the white.graph function. Please can you show me a way of saving the variable names in the gR object?
To export the network to Cytoscape, I used: write.graph(net.result$gR, file = "corr_network_0.7.graphml", format = "graphml") I also noticed the "gml" format returns an error (as mentionned by Guivo) and cannot be opened in Cytoscape. The returned message is as follow: "There were 50 or more warnings (use warnings() to see the first 50)"
Thank you very much for your support, Best regards, Arno