Re: [cytoscape-helpdesk] Plugin to complete node attributes after PSICQUIC Universal Client ?

[Rafael Jimenez]

Hello,

A friend suggested to use ...

http://wiki.cytoscape.org/CyThesaurus_Plugin

The ID mapping might help, but in my opinion it is not a solution to merge networks. Some problems you might encounter ...

1.- Different kind of interactors (proteins, small molecules, genes, ...).

2.- For one gene ID you could find multiple proteins IDs. Gene IDs normally are not specie specific. You could make wrong assumptions with the result of the mapping.

3.- Difficult to make a mapping even with the same type of interactor. i.e. there is not a 100% match between proteins from UniProt and RefSeq, not event between ENSEMBL and UniProt. Moreover the mapping might not be one to one. i.e. One UniProt normally matches several RefSeq.

As you can see in this presentation (slide 15) ... http://www.ebi.ac.uk/~skerrien/talks/2011.04-PSIMI_heidelberg/Service%20Review.ppt ... not all the interactions from all the PSICQUIC can be easily merged/clustered. 

This problem was discussed a couple of years ago, and as result a data distribution best practices document was proposed for data providers ...

http://code.google.com/p/psicquic/wiki/DataDistributionBestPractices

Two main ideas are described in this document to overcome this issue: 1.- Agree on one type of identifiers, 2.- use ROGIDs. It is up to the data providers to apply this best practices and as you realized not all provider do. I hope with time they will be adopted.

I think there is not perfect solution, but the "micluster" algorithm tries its best looking at the different identifiers provided by PSICQUIC ... http://code.google.com/p/micluster/ ... If I am not wrong this merging/clustering functionality was integrated by Kei in Cytoscape 3.

If you are interested to discuss this kind of issues you are very welcome to the April PSI-MI meeting in Liverpool. 

http://www.psidev.info/psi2013

Regards,

Rafael.

On 1 February 2013 15:03, <isabe...@hotmail.com> wrote:

Hi All,

I am trying to use PSICQUIC Universal Client for a set of proteins. I tried to merge the networks obtained from the different databases but it didn't work properly due to IDs differences.
That's why I decided to firstly complete node attributes in the different networks and then to merge them according to a common attribute.

Do you happen to know if there is a special plugin for that ? An 'ID mapping' kind of plugin that could fulfill the missing information ?

Thank you in advance,

Isabelle

--
You received this message because you are subscribed to the Google Groups "cytoscape-helpdesk" group.
To unsubscribe from this group and stop receiving emails from it, send an email to cytoscape-helpd...@googlegroups.com.
To post to this group, send email to cytoscape...@googlegroups.com.
Visit this group at http://groups.google.com/group/cytoscape-helpdesk?hl=en.
For more options, visit https://groups.google.com/groups/opt_out.