[BioC] Limma normalizeBetweenArrays does not normalize M values using Quantile method directly.???

[James W. MacDonald]

Hi Neeraj,

One possible reason for the lack of responses is that you don't appear 
to have closely read the documentation for this method.

>From ?normalizeBetweenArrays:

    Quantile normalization was explored by Yang and Thorne (2003) for
     two-color cDNA arrays. ‘method="quantile"’ ensures that the
     intensities have the same empirical distribution across arrays and
     across channels. ‘method="Aquantile"’ ensures that the A-values
     (average intensities) have the same empirical distribution across
     arrays leaving the M-values (log-ratios) unchanged. These two
     methods are called "q" and "Aq" respectively in Yang and Thorne
     (2003).

I don't see anything there that would indicate that the M-values are 
quantile normalized, do you? It instead appears that the individual 
channels are quantile normalized, unless you specify Aquantile, in 
which the geometric mean intensities are normalized.

Now if we look at the limma User's Guide, under section 6.3, we have:

"This section explores some of the methods available for between-array 
normalization of two-color arrays. A feature which distinguishes most 
of these methods from within-array normalization is the focus on the 
individual red and green intensity values rather than merely on the 
log-ratios. These methods might therefore be called individual channel 
or
separate channel normalization methods."

And a bit later we have:

"Applying quantile normalization directly to the individual red and 
green intensities produces a similar result but is somewhat noisier:
> MA.q <- normalizeBetweenArrays(RG.b, method="quantile")"

So in at least three instances it is clearly specified that quantile 
normalization for two-color arrays is not performed on the M-values, 
and a reference to the rationale behind this process is also provided.

Best,

Jim




On Friday, September 13, 2013 1:17:51 AM, Neeraj Rana wrote:
> Hi dear all,
> I am using limma for two color Agilent Microarray  analysis.
> I found that normalizeBetweenArrays does not perform Quantile normalization
> directly to the M values.
> For that reason i had to use the following command
> MA_Quantile<- noralizeBetweenArrays(*MA$M*,method="quantile")
> To normalize M values i had to give MA$M instead of MA alone.
> I want to make sure whether this way it is  correct or not and why M values
> are not normalized directly while it is mentioned in limma user guide that
> quantile method normalize M values.
>
> This question was raised earlier in Bioconductor list (
> https://stat.ethz.ch/pipermail/bioconductor/2011-May/039618.html) but there
> is no answer for that.
>
> Thanks
>
>
>

--
James W. MacDonald, M.S.
Biostatistician
University of Washington
Environmental and Occupational Health Sciences
4225 Roosevelt Way NE, # 100
Seattle WA 98105-6099