[BioC] Re: affylmGUI

[James Wettenhall]

Pedro,

> I do not know if there is any forum where to ask people or 
> share experience about this package. 

The best place to discuss affylmGUI is on the Bioconductor 
mailing list:
http://www.bioconductor.org/mailList.html
https://www.stat.math.ethz.ch/mailman/listinfo/bioconductor

I hope you don't mind that I've copied my reply to the mailing 
list, in case anyone out there wants to improve upon my answers.

> In the table with ranked genes differentially express what is 
> the column M and A? are the values for the M A plot blot???

The terms M and A are most well known in the two-color 
microarray literature, e.g. see:
http://www.statsci.org/smyth/pubs/mareview.pdf

but they can also be applied to Affymetrix arrays.

M is a log ratio comparing gene expression levels.
  (Base 2 is used for the logarithm)

e.g. log_2(Mutant_expr_level/WildType_expr_level)

A is an average log intensity.
  (Again, the logarithm is in base 2)

Obviously, you can't get absolute gene expression levels from 
microarrays.  For two-color arrays, the A value will depend on 
the scanner intensity.

For two-color arrays it is actually a little confusing because 
we use M to mean a log-ratio between the Red and Green channels 
on one array, but later after we have averaged between replicate 
arrays (or more formally, after we have fitted a linear model), 
then we use M to mean a log-ratio between gene expression levels 
(e.g. comparing Treatment vs Control or Wild-Type vs Mutant).

An M value of positive 1 means a fold-change of 2 and an M value 
of -1 means a fold-change of 0.5 (or equivalently, a fold-change 
of 2 in the other direction).

If you want to convert the M column to fold-changes you can use 
=2^M in Excel, e.g. if the M column is column "E" and the first 
M value was in cell E2, then you could create a column G 
entitled "Fold Change", and in cell G2, you could type the 
formula:
=2^E2
and then fill-down.

(I'm assuming that you know how to get the toptable into Excel 
by first saving as tab-delimited text.)

FDR is False Discovery Rate, a method for multiple-testing 
correction.  I think the default method in 
limma, limmaGUI, affylmGUI is:
Holm, S. (1979). A simple sequentially rejective multiple test
procedure. _Scandinavian Journal of Statistics_, *6*, 65-70.

For references to other multiple-testing correction methods, 
from the R prompt, type:
?p.adjust

Hope this helps,
James

--------------------------------------------------------------------------
James Wettenhall                                  Tel: (+61 3) 9345 2629
Division of Genetics and Bioinformatics           Fax: (+61 3) 9347 0852
The Walter & Eliza Hall Institute         E-mail: wettenhall at wehi.edu.au
 of Medical Research,                     Mobile: (+61 / 0 ) 438 527 921    
1G Royal Parade,
Parkville, Vic 3050, Australia
http://www.wehi.edu.au