how to create a network to be visualized by cytoscape

[jean-char...@univmed.fr]

Hi,
I'm new in cytoscape, and I'm using primarily transcriptomic and
metabolomic data; I tried to use cytoscape to generate network from
some gene expression and metabolomics experiment, but I should say
that the tutorial is of poor use. The data matrix I imported (gene
expression or metabolomic centre-scaled data) did not show any network
at all, altthough no error messages were displayed;
I have few questions:
1) what kind of interactions are used to generate edges? correlations?
is cytoscape generates correlation matrix to be further used for
networing , or is it working like a dendrogram-like generator, or
else? nothing is explained in the tutorial; alternatively, is there a
open source soft or plug-in that I can download that calculates the
interactions to be used in cytoscape (I'm mostly interested in
correlations)?
2) I want to use non-annotated data in cytoscape, just to visualize
how my variables interact together; is that the purpose of cytoscape?
can I download a simple set of data (in a text file format, with no
exotic extensions that no soft can read) with a stepwise explanations
of how to proceed to generate a network?

briefly speaking: I have transcriptomic (microrray, QPCR) and
metabolomic data, and I want to see how my variables (genes,
metabolites) can interact together (either genes-genes, metabolites-
metabolites, or genes-metabolites); can cytoscape do the job? Can I
download a simple data set in a text format that I can inspect to look
how they look like, how they are organised in matrix, and that I can
test to help me? (the data exemple given in cytoscape are not usefull
since I can not open the file extensions).

thanks for help, I deeply need it, and I'm lost with cytoscape

jc

[piet molenaar]

Hi jc,
Cytoscape is an open source software platform for visualizing molecular interaction networks and biological pathways and integrating these networks with annotations, gene expression profiles and other state data. So this means that you usually start with a network of your interest and annotate that with expression data in order to see how your system responds. It is not suited to derive networks automatically from a set of gene-expression data (I think that is basically one of the holy grails of molecular biology anyway; if it only were that simple ;-)
As I understand from what you describe is that you loaded a data matrix; this is loaded as data indeed but in order to visualize it you have to have a network loaded with some type of id for each node that corresponds to ids in the data-matrix.
The basic tutorial explains this mechanism by starting with the creation of example networks and than annotating these with data; have you tried that?
Cheers,
Piet

--
Piet Molenaar
p.mol...@amc.uva.nl
Department of Human Genetics, M1-131
Academic Medical Center
University of Amsterdam
Meibergdreef 9
1105 AZ Amsterdam
the Netherlands

tel (+31) 20-5666592
fax (+31) 20-6918626

[jean-charles.martin]

Hi  Piet

 

Thanks for replying ; in fact, since I posted my add on the help desk, I succeeded to find what I was looking for, that is to visualize a correlation network generated by cytoscape; the thing is that I did not download the appropriate plugin at that time, which answer exactly to my problem (expression correlation network plugin, found in the last cytoscape version); so I’m fine, and happy!!

 

Cheers

 

jc

 

De : cytoscape...@googlegroups.com [mailto:cytoscape...@googlegroups.com] De la part de piet molenaar
Envoyé : vendredi 5 décembre 2008 10:32
À : cytoscape...@googlegroups.com
Objet : [cytoscape-helpdesk] Re: how to create a network to be visualized by cytoscape

[jitesh dundas]

Dear Sir/Madam,

I am looking for some work in this project.

I am working as a research associate( boinformatics and computer science ) and maybe you could help me in the same.

Is there anybody in need of help..

Thanks &  Regards,
Jitesh Dundas
http://jiteshbdundas.blogspot.com

[zwangwustl]

You have any idea how the network is constructed by the plugin? Is
that constructing based on the Pearson's correlation coeffiency? Does
it distinguish between negative and positive correlations?

Best

On Dec 5 2008, 3:46 am, "jean-charles.martin" <jean-

> p.molen...@amc.uva.nl

[jean-charles.martin]

You might construct your hown network 'manually'in file/new/empty network or
construct network from cpath, import it from the web, or generate ones from
expression results sorted in a matrix where observations are in the first
column and variables in rows (use the import function to download), then you
might calculate a correlation network (pearson correlation) using the plugin
'expression correlation network'; you can then visualize the network (number
of nodes and edges depending on the correlation threshold you choose) using
the layout functions.
It's ok with some practice...
Also, you might visualize the positive/negative correlation using the
advance function in vizmapper. To look how data look likes to generate a
network, you might open a *.sif file (network) using the notepad or xcel
(even better). Have a look on the help too.
Good luck!
Hope its helpfull

-----Message d'origine-----
De : cytoscape...@googlegroups.com
[mailto:cytoscape...@googlegroups.com] De la part de zwangwustl
Envoyé : vendredi 2 janvier 2009 22:08
À : cytoscape-helpdesk

[zwangwustl]

Thanks, Jean, for the reply,

Do you do know the methodology that the "correlation expression"
plugin uses to construct the network? Is it based on Pearson's
correlation coefficiency? And how does it distinguish positive and
negative correlations?

Best


On Jan 3, 2:59 am, "jean-charles.martin" <jean-

[piet molenaar]

Find here http://baderlab.org/Software/ExpressionCorrelation an explanation of the methodology by the group that developed the plugin.
Hope this helps,
Piet

--
Piet Molenaar
p.mol...@amc.uva.nl