[BioC] normalization of a data frame

[Henrik Bengtsson]

Hi,

one option is to use the normalization methods in 'aroma.light', which
take matrices (or lists) containing signals on the intensity scale
(non-logged signals), so you could do:

 library(aroma.light)
 X <- as.matrix(df)

and pick one of:

 Xn <- normalizeAverage(X)
 Xn <- normalizeQuantile(X)
 Xn <- normalizeCurveFit(X)

followed by

 dfN <- as.data.frame(Xn)

For normalizeCurveFit() you can specify either method="loess"
(default/slowest), "lowess", "spline", or "robustSpline".  See
help(aroma.light) for more details and help(normalizeCurveFit.matrix)
for an example.

Hope this helps

/Henrik

PS. I'll add it to the to do list for the next release to add
normalization wrappers for data.frame:s as well so you don't have to
do the above coercions. DS.


On Feb 11, 2008 11:26 PM, joseph <jdsandjd at yahoo.com> wrote:
>
>
> Hi
>
> I have a data frame that looks like the one below.  I would like to apply to its columns the different
> kinds of normalizations such as median global normalization (Scaling), Quantile
> Normalization, and Lowess. Do I have to use packages such as affy, Limma and
> marray and if so what do I have to do to the data frame in order to be able to
> analyze it with these packages?
>
>
> Thanks
>
>
>
>
>
> df=data.frame (row.names=LETTERS[1:10], col1=c(1,4,7:9,2:6),
> col2= c(1:4,9, 10:14), col3=c(14,6:10,8:11))
>
>
> > df
>
>
>   col1 col2 col3
>
>
> A    1    1
> 14
>
>
> B    4    2
> 6
>
>
> C    7    3
> 7
>
>
> D    8    4
> 8
>
>
> E    9    9
> 9
>
>
> F    2   10
> 10
>
>
> G    3   11
> 8
>
>
> H    4
> 12    9
>
>
> I    5   13
> 10
>
>
> J    6   14
> 11
>
>
>
>
>
>
>
>
>
>
>
>
>       ____________________________________________________________________________________
> Looking for last minute shopping deals?
>
>         [[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
>