[BioC] MAGE-ML reading

[Sean Davis]

On Apr 1, 2005, at 3:34 PM, Vincent Carey 525-2265 wrote:

>
>> I would like to read MAGE-ML, but I have not been able to get SJava to
>> install on MacOS.  Are there other options to read MAGE-ML besides
>> RMAGEML (which requires SJava)?
>
> There is a version of SJava that ports to Mac OSX.  It is made
> available by Simon Urbanek
>
> http://stats.math.uni-augsburg.de/~urbanek/SJava_0.65-X.tar.gz
>
> I have not tried it with R 2.0.1.  This version of SJava
> lacks some important features of SJava_0.67-3 which is known
> to work with RMAGEML.  You may have to modify the RMAGEML
> source a bit to get it to install with this old version of
> SJava -- you may have to relax the SJava version dependency.
> And this may lead to problems.  But I think it is worth a
> try.
>
> You could parse and process the MAGE-ML using the XML package.
> If you had a clear and fairly restricted set of elements that you
> needed to grab, this might be very effective.
>
> The nice thing about RMAGEML is that it can in principle
> reflect anything in the MAGE-OM as captured by the java MAGEstk.
> RMAGEML has the resources to translate elements of MAGE-OM directly
> into elements of marray class instances.
>
> There is also a perl MAGEstk, I believe.  One might get an
> interface through RSPerl.
>
> Please keep us informed of successes and failures in using
> MAGEML documents with Bioconductor.
>

Thanks, Vince.

Just so you know why I ask, one of our new array platforms dumps 
straight MAGE-ML, so I am excited to have those data available.

As for RSPerl,  I haven't had much luck (although I haven't tried with 
2.0.1) with installing it, either.  I'll need to give it a try again, 
as I do use perl quite a bit.

Sean