Hi Kaushal,
Please keep your posts on the list so you may get help from others as
well.
It is not clear to me how your data were normalized in beadstudio, but
typically it is not a good idea to normalize your data in beadstudio.
The neqc function in limma does a much better job for this.
I can see from your code that your design matrix was not generated
properly. Have a look at Illumina case study (section 15.3, "Comparing
Mammary Progenitor Cell Populations with Illumina BeadChips") in limma
users guide, which should be helpful for your data analysis. Type
'limmaUsersGuide()' at R prompt will bring up the users guide. Please
also provide session info (output of 'sessionInfo' command) when your
report problems so that we can know which versions of packages you
use.
Best wishes,
Wei
On Sep 5, 2013, at 12:46 AM, Kaushal Chaudhary wrote:
> Dear Dr. Shi,
>
> Data was already normalized from beadstudio. Here is the small part
of data set and the code. I have also mentioned what type of error
occured. I tried to create design matrix for the data to use "limma"
package. It is not giving me result. I really appreciate your help
in this regard. Thank you.
>
> Sample dataset
>
> ProbeID
> mt
> mt
> mt
> wt
> wt
> wt
> 2600193
> 7.388713
> 7.422518
> 7.340093
> 7.531799
> 7.260194
> 7.374177
> 2370397
> 7.496064
> 7.490054
> 7.215325
> 7.399284
> 7.721715
> 7.277656
> 1710328
> 7.200958
> 7.277063
> 7.265149
> 7.116158
> 7.272333
> 7.120288
> 3450193
> 7.378917
> 7.547722
> 7.393085
> 7.292157
> 7.189327
> 7.475774
> 1570300
> 7.755082
> 7.634331
> 7.652033
> 7.84686
> 7.428635
> 7.942549
>
>
>
> data1=as.matrix(read.csv("kamesh.csv", header=T))
>
> data2=log(data1)
>
> dim(data2)
>
> class(data2)
>
> data3=data2[,-1]
>
> head(data3)
>
> source("
http://bioconductor.org/biocLite.R")
>
> biocLite("limma")
>
> hist(data3)
>
> boxplot(data3)
>
> xdist <- dist(t(data3))
>
> dim(t(data3))
>
> hc <- hclust(xdist)
>
> plot(as.dendrogram(hc))
>
> image(data2)
>
>
> library("gplots")
>
> heatmap.2(data3)
>
>
> ###Error: cannot allocate vector of size 7.6 gb
>
>
> mat<-matrix(0,45281,6) ### creating design matrix 45281 is row of
genes and 6 samples on the columns (mt,mt, mt, wt,wt,wt)
>
> mat[,c(1,2,3)]<-1
>
> mat
>
> library("limma")
>
> fit=lmFit(data3, mat)
>
> #### Coefficients not estimable: 2 3 4 5 6
>
> Error in lm.fit(design, t(M)) : incompatible dimensions
>
>
>
>
> On Tue, Sep 3, 2013 at 10:12 PM, Wei Shi <shi@wehi.edu.au> wrote:
> Dear KaushalRaj,
>
> The heatmap.2 function in gplots package is a useful tool for making
heatmaps. But I can't see how you normalized your data from your
provided code. The neqc function in limma can be used to background
correct and normalize your illumina data.
>
> Hope this helps.
>
> Cheers,
> Wei
>
> On Sep 4, 2013, at 12:23 PM, KaushalRaj Chaudhary [guest] wrote:
>
> >
> > Hi,
> > I have a one color illumina microarray dataset with six samples (3
wild type and 3 mutant types ) on the columns and around 48000 genes
on the rows of excel. I want to get the hierarchical clustering of
genes but I am getting clustering of samples only. Can I use limma
package for the analysis of this data and also get heatmap. The data
is already normalized. Thank you very much for your help.
> >
> > Regards,
> >
> > -- output of sessionInfo():
> >
> > data1=as.matrix(read.csv("kamesh.csv", header=T))
> > data2=log(data1)
> > dim(data2)
> > class(data2)
> > data3=data2[,-1]
> > head(data3)
> > source("
http://bioconductor.org/biocLite.R")
> > biocLite()
> > biocLite("limma")
> > biocLite("lumi")
> >
> > hist(data3)
> > boxplot(data3)
> > xdist <- dist(t(data3))
> > dim(t(data3))
> > hc <- hclust(xdist)
> > plot(as.dendrogram(hc))
> > image(data2)
> >
> >
> >
> > --
> > Sent via the guest posting facility at bioconductor.org.
> >
> > _______________________________________________
> > Bioconductor mailing list
> > Bioconductor@r-project.org
> >
https://stat.ethz.ch/mailman/listinfo/bioconductor
> > Search the archives:
http://news.gmane.org/gmane.science.biology.informatics.conductor
>
>
______________________________________________________________________
> The information in this email is confidential and
inte...{{dropped:17}}
