[BioC] Agilent MAGE-ML

Martin Morgan mtmorgan at fhcrc.org
Wed Aug 20 18:34:22 CEST 2008


Hi Gabriele --

A little slow on the response, but I wanted to provide a couple of
not-very-complete suggestions. 

gabriele bucci <gabriele.bucci at ifom-ieo-campus.it> writes:

> Dear List,
> some external collaborators gave us the results of an expression
> profiling over tumor samples.
> The experiment is a dye swap, and they are using a rosetta resolver
> platform.
> The chips used were Agilent two-colors custom arrays and the data
> which were provided to us are in the form of ".mage" files.
> So I suppose these files are actually a MAGE-ML files.
>
> Except for the package 'RMAGEML' which run only under a Linux R
> environment (I'm working under winXP), I was not able to find a
> Bioconductor solution to read and process this MAGE-ML files.

'AffyCompatible' will parse MAGE files; unfortunately, it requires a
version of the XML package (e.g., 1.95-3) that is different from the
one available by default. Try

> install.packages('XML', repos='http://www.omegahat.org/R')
> source('http://bioconductor.org/biocLite.R')
> biocLite('AffyCompatible')
> library(AffyCompatible)
> example(readMage)

You'll end up with a complicated object that you can traverse (use the
names suggested by the display of the object, with vector-like
elements suggested by () and accessible with [[) to retrieve relevant
information; there is no easy way to convert this to, say, an
ExpressionSet, and the MAGE model is horrendously complicated.

http://www.ebi.ac.uk/microarray/doc/software/schema/MAGE/MAGE.htm

The help page for readMage also points to readXml, which will be a
more efficient way to retrieve information once you know what it is
that you want. There are useful vignettes in the package, too.

A further complication is that Affy used MAGE for sample attribute
description, not for storing expression data; Rosetta Resolver might
well do the same thing but I have no direct experience. Also MAGE
implementations are not always compliant with the standard, so the
MAGE in AffyCompatible may differ in obscure ways from the MAGE used
by Rosetta.

> My questions are:
>
> 1) when talking in general about 'Agilent .txt files'  we`re actually
> talking about MAGE files?

If you have Agilent .txt files, then probably this is the data that
will be most useful in the analysis, e.g., using limma's
read.maimages.

> 2) is there a way to port RMAGEML on R under windows?
> 3) alternative solutions.
>
> I will appreciate any kind of suggestion and comment, if you have some.
>
> My Best wishes,
> Gabriele
>
> PS. I've tried also with Partek and GeneSpring GX (7.3) without success...
>
> -- 
> Gabriele Bucci, PhD 
>
> COGENTECH  -Consortium for Genomic Technologies- 
> Via Adamello 16 20139 Milano  -Italia- 
> Tel.    +39 02 94375134 Fax.    +39 02 94375991
> gabriele.bucci at ifom-ieo-campus.it
> service-array at ifom-ieo-campus.it
>
> _______________________________________________
> Bioconductor mailing list
> Bioconductor at stat.math.ethz.ch
> https://stat.ethz.ch/mailman/listinfo/bioconductor
> Search the archives: http://news.gmane.org/gmane.science.biology.informatics.conductor

-- 
Martin Morgan
Computational Biology / Fred Hutchinson Cancer Research Center
1100 Fairview Ave. N.
PO Box 19024 Seattle, WA 98109

Location: Arnold Building M2 B169
Phone: (206) 667-2793



More information about the Bioconductor mailing list