Network visualization – tutorial, Jan 29, 2009 - BiGRe


1 Οκτ 2013 (πριν από 3 χρόνια και 6 μήνες)

74 εμφανίσεις


BioSapiens European School of Bioinformatics

Network visualization

tutorial, Jan 29, 2009

By Syvain Brohée, BiGRe laboratory, ULB


To help scientists apprehending their network of interest, the first step is to visualize it in a graphical

Networks are generally displayed as a set of dots (or boxes) representing nodes, linked by lines (edges) or

(arcs in the case of directed graphs). Nodes and the edges may be associated to a label and/or a
weight. Node labels generally appear i
n the node boxes, whereas edge labels are often placed above the

In this short tutorial, we will introduce, CytoScape, the major tool for network visualization.


is one of the most popular bioinformatics software dedicated to the study of b
networks as it combines many advantages.


It is an open source software that supports network visualization (among other features). Indeed, it
includes a set of
layout algorithms

that permit to adapt the node positions according to various
s, in order to better detach them for each other, and highlight their relationships.


In its basic version, CytoScape is mainly devoted to network visualization. However, one of its major
strength is the availability of a large number of
in modules

itten by different developers, in
order to provide CytoScape with additional capabilities (clustering, annotation, topological analysis


CytoScape is written in Java, so that it runs on all platforms (Mac OS X, Unix, Windows).

CytoScape and its plug
ins can be downloaded from





Another tool that might be interesting for network visualization is
a free Java tool almost only
dedicated to network visualization.
Some of the yEd layout functions (e.g. layout algorithms) are based
on the same source code that CytoScape, but they support the specification of more parameters. yEd is
freely downloadable from

Protocol for CytoScape


With a Web browser, open a connection to the supplementary material of this course



click on the file string co
expression file and download it to your computer.


This demonstration graph consist

in the top scoring edges of the yeast co
expression network included in
the integrative database

Jensen et al, 2009
). It contains 537 nodes and 4801 edges. In this network,
nodes represent genes
, and the presence of a weighted edge between two genes means that these two
genes were found to be co
expressed in some microarray experiment. Note that each edge is weighted
the reliability of the

detected co

The network is in the G
ML format, for a detailed description of graph formats, refer to the protocol in
Brohée, Faust et al (2008).



On the computers of the class room, type



Import the network you downloaded in CytoScape.

In the CytoScap
e menu, select the command



Network (Multiple File Types)

Select the file and click
. After the importation, the left panel indicates the name of the
network, as well as the number of nodes (537) and edges (4801).

The network is d
isplayed, but its layout makes it very uncomfortable to interpret: nodes are simply
aligned along a diagonal). In the next steps, we will see how to apply a more readable layout.

To zoom in and zoom out, use the roll of your mouse or right click and move
upwards (zoom in)
or downwards (zoom out). To see the entire network in the window, click on the “1:1 in a
magnifying glass” button.


You can now play with the different layout algorithms of CytoScape. These are available under the
layout menu. I would enco
urage you to mainly use the following layouts


Organic / Spring
Embedded / Force
This layout is based on an analogy with the
physical forces of springs submitted to some tension. Nodes are considered as physical objects
with mutually repulsive for
ces (like, e.g., protons or electrons). The connections between nodes
also follow the physical analogy and are considered to be metal springs attached to the pair of
nodes. These springs produce repulsive or attractive forces between their end points if th
ey are too
short or too long. The

simulates these physical forces and rearranges the positions of the
nodes in such a way that the sum of the forces emitted by the nodes and the edges reaches a (local)

This is one of the most frequently
used layout in bioinformatics as it provides an
intuitive visualization of groups of nodes that are tightly joined.


Circular layout

group and tree structures within a network. It creates node partitions
by analyzing the connectivity structure of

the network, and arranges the partitions as separate
circles. The circles themselves are arranged in a radial tree layout fashion.

Now that the network is “unfolded”, you can observe that the edges present different colors and width.
Edge widths are propo
rtional to their weights.


To visualize the label or the weight of the network, click vizmapper in the left frame of CytoScape.
Under the “
Unused properties
” section, double click on Edge Label. Edge label is now in the “
Visual Mapping
” section, selec
t “
Passthrough mapping
” as the Mapping Type.

On each edge, the weight or the label of the edge is now specified (with the names of the two
nodes linked by the edge).

Protocol for yED


Open yED

On the computers of the class room, type



Open the

demonstration file with
File > Import


Change the layout with the Layout menu.

You can observe that there are more options that with the CytoScape layout. For example, for the
spring embedding layout, yEd allows you to specify several parameters, such as

the attraction
between two nodes linked by an edge or the repulsion of two nodes. You can also apply the layout
only to a subset of selected nodes.


Brohée S, Faust K, Lima
Mendez G, Vanderstocken G, van Helden J., 2008.
Network Analysis Tools:
from biological networks to clusters and pathways. Nat Protoc. 3(10):1616

Jensen LJ, Kuhn M, Stark M, Chaffron S, Creevey C, Muller J, Doerks T, Julien P, Roth A,
Simonovic M, Bork P, von Mering C. STRING 8
a global view on proteins and their funct
interactions in 630 organisms.
Nucleic Acids Res. Jan;37 (Database issue):D412

Shannon P, Markiel A, Ozier O, Baliga NS, Wang J., Ramage D, Amin N, Schwikowski B, and
Ideker T, Nov 2003. CytoScape : a software environment for integrated mode
ls of biomolecular
interaction networks. Genome Res, 13(11) :2498