[BioC] affy cel + cdf files content

[Laurent Gautier]

On Mon, Jul 01, 2002 at 09:52:04AM -0400, Robert Gentleman wrote:
> On Mon, Jul 01, 2002 at 02:39:31PM +0200, Laurent Gautier wrote:
> > On Fri, Jun 28, 2002 at 08:23:07AM +0200, julien.sylvestre@wotan.ens.fr wrote:
> > > This message is not directly related to Bioconductor yet I hope it might be of general interest for users...
> > > 
> > > The first line of an Affymetrix Yg_s98 yeast chip .cel file looks like this
> > > X	Y	MEAN	STDV	NPIXELS
> > > 0	0	232	45	36
> > > I read about 100 pixels were used whereas it seems that no more than 36 are used here. 
> > > How is the svd calculated ?
> > 
> 
>   The mean is actually the 75th percentile. I believe that the STDV is 
>  actually the standard deviation (as defined in any stats book). Both
>  are based on the pixels used as I recall. But in all cases, one
>  really has to refer to Affymetrix documentation. This is generally
>  supplied when you buy chips and or a machine. Some of it is available
>  on their website www.affymetrix.com.
> 
>    Robert
> > 

Thanks for the clear answer Robert. In haste I only generated almost random
blurb.

I remembered observing lines like below in the header of CEL files

---------------
Algorithm=Percentile
AlgorithmParameters=Percentile:75;CellMargin:2;OutlierHigh:1.500;OutlierLow:1.004
--------------

...I do not know if there is a way to change the algorithm/parameters in MAS...





L.