[BioC] question about scale factor for Affy chip

[Hooiveld, Guido]

It is A, since you scale all arrays to the same target value (100 is
default in BioC, 500 is default in GCOS).

According to Affymetrix'criteria the difference between the SFs should
not vary more than a factor of 3. Since in your case the SFs are within
the same range (10.-10.2), the SF does not indicate a problem. If you
would have had 10.2. 10.1, 10.4, 40.9 and 2.7, the SF indiactes the last
two arrays deviate from the other arrays.

HTH,
Guido
 

> -----Original Message-----
> From: bioconductor-bounces at stat.math.ethz.ch 
> [mailto:bioconductor-bounces at stat.math.ethz.ch] On Behalf Of 
> James Anderson
> Sent: 29 June 2007 16:32
> To: bioconductor
> Subject: [BioC] question about scale factor for Affy chip
> 
> Hi, 
> 
> I am using the R package SimpleAffy, now I have a conceptual 
> question regarding the scale factor. Seems that Affymetrix 
> recommend the scale factor should be within 3 fold, the way I 
> understand the scale factor is that it is the ratio between 
> the trimmed mean of one chip (normalized by MAS5) to the 
> target value (say, 100 or 500). According to the manual for 
> QC in SimpleAffy: 
> 
> "Large variations in scale factors signal cases where the 
> normalisation assumptions are likely to fail due to issues 
> with sample quality or amount of starting material."
> 
> I am getting a little confused about EXACTLY what does scale 
> factor mean:
> A. The ratio between the trimmed mean of each chip to a 
> target value B. The ratio between the trimmed mean of each 
> chip to the trimmed mean of all the chips. 
> 
> because the above section seems to me that as long as there 
> is small variation in the scale factor, the quality is no 
> problem. What about all the arrays have scale factor in the 
> range (10.1-10.2)? The scale factor has almost no variation, 
> but they are much higher than 3. or I am misunderstanding some part? 
> 
> 
> Thanks a lot!
> 
> James
> 
>        
> ---------------------------------
> Pinpoint customers who are looking for what you sell. 
> 	[[alternative HTML version deleted]]
> 
> _______________________________________________
> 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
>