[BioC] GEOquery doubt: converting GDS into a MAlist

[Sean Davis]

On Thu, Apr 24, 2008 at 9:52 AM, Juan C Oliveros Collazos
<oliveros at cnb.csic.es> wrote:
> Hi all,
>
>  I am trying to dowload a set of data samples from GEO database by using
>  GEOquery package:
>
>  Example:
>  -------------------------------
>
>  library (GEOquery);
>
>  gds <- getGEO("GDS1047");
>
>  gpl<-getGEO("GPL198");
>
>  MA <- GDS2MA(gds, GPL = gpl);
>  --------------------------------
>
>  By the GEOquery user manual, this result in a MAlist containing all
>  details about the experiment.
>
>  As far as I understand, the MAlist object is used for storing M values
>  and A values, amongst other details.
>
>  * M values are defined as the logRatio of all genes in each comparison.
>
>  * A values are defined as the average of logIntensities for each comparison.
>
>  However, the MAlist created with GDS2MA seems not to store in "M" the
>  logRatios but only the intensities corresponding to every GSM in the
>  dataset.
>
>
>  I need to use the logRatios (M) of a lot of experiments and I would
>  write an script that automatically download and store these M values in
>  tables.
>
>  Obviously, the script should be able to deduce which comparisons are
>  involved in the dataset (which GSM are numerators and which are
>  denominators)
>
>  Now I can store the intensities of the GSMs, but to convert these
>  intensities in ratios I need to check which are the comparisons by
>  hand... I would like to avoid this.

Hi, Juan.  GDS1047 is a single-channel experiment.  Therefore, there
are not ratios and GEOquery can't know how to make such ratios.  As
for downloading and determing pairings for making ratios, you will, in
general need to do that by hand.

Hope that helps.

Sean