[BioC] limma matrix design for biological replicates.

[Gordon K Smyth]

Dear Neeraj Rana,

You targets file is ok, although your samples are really normal and tumour 
rather than wt and mutant.  You can use

   design <- model.matrix(~factor(targets$Cy5))
   colnames(design) <- c("NodNeg","PosvsNeg")
   ...
   topTable(fit, coef="PosvsNeg")

Best wishes
Gordon


> Date: Fri, 25 Jun 2010 13:55:46 +0530
> From: neeraj <kushrn at gmail.com>
> To: Bioconductor at stat.math.ethz.ch
> Subject: [BioC] limma matrix design for biological replicates.
>
> hi,
>
> i am doing analysis for 10 breast cancer arrays (two color Agilent4x44
> arrays) with limma,where Cy3 is normal and Cy5 is tumor.And all are
> biological replicates ,means all 10 arrays have been prepared from 10
> different patients.
> there are two categories as nod negative and nod positive.Five arrays from
> nod negative(5 patients) and five are from nod positive(another 5
> patients).I want to see the differentially regulated genes between two
> categories(NOD NEGATIVE vs NOD POSITIVE).
> i designed the target file as given below..I want to make it sure whether it
> is correct or not.
>
> SampleNumber                 FileName         Cy3              Cy5
> 1                                      1135_NN.txt
> wt1                 mu1
> 2                                      2157_NN.txt              wt1
>        mu1
> 3                                      3159_NN.txt
> wt1                 mu1
> 4                                      4171_NN.txt
> wt1                 mu1
> 5                                      5179_NN.txt              wt1
>                mu1
> 6                                      628_NP.txt
> wt2                 mu2
> 7                                      758_NP.txt
> wt2                 mu2
> 8                                      880_NP.txt
> wt2                 mu2
> 9                                      993_NP.txt
> wt2                 mu2
> 10                                   1096_NP.txt
> wt2                 mu2
>
> 1 to 5 are nod negative and ,6 to 10 are nod positive.
>
> thanx.
> NEERAJ RANA
> JRF
> IISC BANGLORE(INDIA)

______________________________________________________________________
The information in this email is confidential and intend...{{dropped:4}}