Re: getting normalized expression values from GEO GSE files

On Wed, Aug 27, 2014 at 4:18 PM, Maria Kesa <[email protected]> wrote:

> 2. How should I normalize the data, considering that there are multiple
> platforms in the experiment?
>

It's not obvious from the GEO page or the paper why they used 7 platforms,
but I believe they may be complementary and intended to be combined to
provide something approaching whole-genome coverage.  I don't envy you
trying to normalize these spotted cDNA arrays, but I would do a standard
normalization such as Loess, and do exploratory analysis such as seeing
whether the different platforms have different batch effects (could be
apparent if some platforms have many more differentially expressed genes,
for example, or very different sample clustering patterns than other
platforms).

	[[alternative HTML version deleted]]

_______________________________________________
Bioconductor mailing list
[email protected]
https://stat.ethz.ch/mailman/listinfo/bioconductor
Search the archives: http://news.gmane.org/gmane.science.biology.informatics.conductor