[R] principal component analysis in affy

[Adaikalavan Ramasamy]

1) Please learn to wrap your emails to 72 characters per line. See 
http://expita.com/nomime.html

2) You might have better luck with the BioConductor folks. Their mailing
list is https://stat.ethz.ch/mailman/listinfo/bioconductor

3) The affy package has many functions including some algorithms for
preprocessing CEL files.

4) I do not understand why you need the affy package if you do not have
CEL files ? I presume that you only have a subset of the final data.

5) I do not understand your question but you might want to simply
transpose the matrix before doing a prcomp() on it. See help("t").

Regards, Adai



On Thu, 2005-07-21 at 10:09 -0500, Wagle, Mugdha wrote:
> Hi,
>  
> I have been using the prcomp function to perform PCA on my example microarray data, (stored in metric text files) which looks like this:
>  
>         1a 1b 1c 1d 1e 1f ...................................................4r 4s 4t
> g1    1.2705 1.2766 ...........................................................2.0298
> g2    0.1631 ........................................................................0.7067
> g3    0.2212 ........................................................................1.0439
> .
> .
> .
> .
> g99  1.3657..........................................................................2.3736
>  
> i.e. a matrix of 63 columns and 99 rows, where the columns represent chip and rows represent genes. Now, the biplot function
>  
> biplot(prcomp(pcadata, scale = TRUE), cex = c(0.75,0.75))
>  
> gives me a plot with one vector per gene. However, I actually need to get one vector per chip instead of one vector per gene.  I have been told that there is a function in the affy package that does what I am looking for i.e. gives one vector per chip. Can someone please tell me what the function is called, and how I can get hold of the code(since I believe affy only works on CEL files) ? I have downloaded the affy R code from Terry Speed's website already, but I don't know where (if at all) the code to perform PCA is.
>  
> Thank you everyone!
>  
> Sincerely,
> Mugdha Wagle
> Hartwell Center for Bioinformatics and Biotechnology,
> St.Jude Children's Research Hospital, Memphis TN 38105
> 
> ______________________________________________
> R-help at stat.math.ethz.ch mailing list
> https://stat.ethz.ch/mailman/listinfo/r-help
> PLEASE do read the posting guide! http://www.R-project.org/posting-guide.html
>