Control Cytoscape from R Console

Introduction

CytoscapeRPC is a plugin for Cytoscape to use Cytoscape as a RPC server. You can call Cytoscape functions from variety of languages including Perl, Python, Ruby, and R. In this tutorial, you learn how to use Cytoscape from R console.

Setup

You need to install the following plugins/packages.

Cytoscape Plugin

Just copy the following jar files to Cytoscape plugin directory.

• CytoscapeRPC Plugin 0.95 or later.

• Apache XML RPC library

  • Required files are in lib directory: xmlrpc-common-3.1.3.jar, xmlrpc-server-3.1.3.jar, ws-commons-util-1.0.2.jar

R Packages

• XML-RPC package XMLRPC_0.2-mod.zip (Bug fix version by Paul Shannon. BSD License)

• igraph graph utility package

Tutorial

Start Cytoscape RPC

1. Start Cytoscape. If you want to see the all messages (including Exceptions from CytoscapeRPC plugin), start Cytoscape from command line (by executing cytoscape.sh or cytoscape.bat)

2. Select Plugins-->CytoscapeRPC-->Activate RPC

3. Modify port number if necessary. Default port is 9000.

Start R

1. Start R console
2. Load required libraries

   • > library(igraph)
     > library(XMLRPC)

3. Test RPC Plugin.

   • > testResponse = xml.rpc('http://localhost:9000', 'Cytoscape.test')
     > testResponse
     [1] "It works!"

Create Network with igraph and Visualize It in Cytoscape

1. Create an empty network.

   • > networkID = xml.rpc('http://localhost:9000', 'Cytoscape.createNetwork', 'R-Cytoscape Test')
     > networkID
     [1] "2"
     > networkTitle = xml.rpc('http://localhost:9000', 'Cytoscape.getNetworkTitle', networkID)
     > networkTitle
     [1] "R-Cytoscape Test"

2. Generate a scale-free graph according to the Barabasi-Albert model

   • > g1 <- barabasi.game(200)
     > g1
     Vertices: 200 
     Edges: 199 
     Directed: TRUE 
     Edges:
                     
     [0]     1 ->   0
     [1]     2 ->   0
     [2]     3 ->   1
     .
     .
     .

3. Create Cytoscape network from the generated graph

   • > edgelist1 <- get.edgelist(g1)
     > edgelist1
            [,1] [,2]
       [1,]    1    0
       [2,]    2    0
       [3,]    3    1
     .
     .
     .
     > edgeIDs <- xml.rpc('http://localhost:9000', 'Cytoscape.createEdgesFromVector', edgelist1[,1], edgelist1[,2])

4. In the Cytoscape window, you can see one red node. Actually, it's a graph with 200 nodes without any visual properties. After applying layout and vizmap, you can see something like this:

   • 

(to be continued...)