[BioC] background correction with cDNA array data

[Sean Davis]

On 10/26/05 11:01 PM, "Qi.Zhang at astrazeneca.com" <Qi.Zhang at astrazeneca.com>
wrote:

> Naomi, thanks for your reply.  I do not intent to write my own mathod.  I
> took the default option when read in the gpr file using read.GenePix, so I
> assume the background correction was done by "substraction".  I remember it
> from somewhere that the "substraction" may introduce additional variance
> depends on the algorithm of the background estimation (from the iamge
> analysis software).  Is the "substraction" the only bg correction method in
> marray?  limma has more options, but I have no idea about the advantage or
> disadvantage of them.  Any suggestions will be very helpful.

All the bioconductor packages contain extensive help, and limma has some of
the best help files available.  Simply reading the help for the
backgroundCorrect function in limma gives you a rather concise description
of the various options available.  There are also references given which you
can certainly pull for more information.  If contemplating using limma, a
required read is the relevant section(s) of the limma user guide.  All that
said, there is nothing wrong with trying each method and then plotting your
data to get a feel for what is really going on with your data.

Sean