[BioC] CQN Package

[Kasper Daniel Hansen]

On Mon, Jun 25, 2012 at 9:46 AM, Xin Davis <xydavis at ncsu.edu> wrote:
> Dear All,
>
> I try to use CQN package to normalize RNAseq data for modeling, but
> could not find CQN vignette at Bioconductor website.
>
> The example from the manual is not clear to me (shown below). I should be
> able to run the code by replacing montgomery.subset with our dataset, but
> it is not the case.
>
> I assume montgomery.subset is the data set, what is sizeFactors.subset ?
> Other pacakge (DESeq, edgeR) will calculate sizeFactors. How about uCovar ?
> I Should provide dataset, the package will calculate whatever required by
> the package. The explanations are not clear to me.

There are many suggested ways to estimate sizeFactors.  You have the
option of inputting whatever you want.  The default will use
  colSums(data)

You need to give the function information about the gc content and the
gene lengths.  uCovar in the vignette is a matrix with 2 columns (gc
content and gene length).  What these values are, depend entirely on
how you got your count matrix.   The uCovar in the package fits
together with how the data in the package was computed from the
aligned reads, but of course it will most likely not fit with your
data.

If you want to, you can inspect these objects yourself by loading the package.

Kasper


>
> I would appreciate it if anyone provide guidance on this.
>
> Thanks,
> Xin Davis
>
>
>
>
> data(montgomery.subset)
> data(sizeFactors.subset)
> data(uCovar)
> cqn.subset <- cqn(montgomery.subset, lengths = uCovar$length,
> x = uCovar$gccontent, sizeFactors = sizeFactors.subset, verbose = TRUE)
>
>        [[alternative HTML version deleted]]
>
> _______________________________________________
> Bioconductor mailing list
> Bioconductor at r-project.org
> https://stat.ethz.ch/mailman/listinfo/bioconductor
> Search the archives: http://news.gmane.org/gmane.science.biology.informatics.conductor