[BioC] NCI pathways to graphNEL

[Gilbert Feng]

Hi, Laurent

Basically, GeneAnswers is designed for enrichment test based on 
different annotation libraries. In your case, if you just want to 
retrieve one or several NCI pathways, you can use standard xml query to 
obtain them from caBIO site. Then reconstruct the pathway by yourself 
(GeneAnswers also contains filtered gene/protein interaction dataframes 
for human, mouse and rat from NCBI). But if you want to retrieve a lot 
of NCI pathways, xml query might take a long time and it's better to ask 
NCI or caBIO people to generate a flat text for you. Also, caBIO doesn't 
includes only NCI, but also old BioCarta and Reactome pathways.

Best

Gilbert



On 2/3/11 11:34 AM, laurent jacob wrote:
> Hello,
>
> I'm trying to import NCI pathways (available in biopax format) to graphNEL
> objects.
>
> This question was already asked on the list but with no fully satisfactory
> answer (for my purpose):
>
> - Some pointed to Rredland, which I tried and it indeed seems to parse the
> .owl properly, but then the conversion to graphNEL is not straightforward.
>
> - GeneAnswers was also mentioned, but from what I saw in the documentation,
> it only loads the gene lists, not the graph of the interactions between the
> genes in the pathway.
>
> - Finally, somebody proposed to use the sif format, which would be fine for
> a very simplified DAG, but I would need to keep the type of interactions
> too.
>
> Has anyone solved the problem since the last thread?
>
> Thanks,
>
> Laurent<http://cbio.ensmp.fr/%7Eljacob>
>

-- 
-----------------------------------------------
Gang (Gilbert) Feng, PhD
Biomedical Informatics Center
Robert H. Lurie Comprehensive Cancer Center
Northwestern University
750 N. Lake Shore Drive, 11th Floor(11-175e)
Chicago, IL  60611
Phone:312-503-2358
Email g-feng (at) northwestern.edu