Dear all,
I am transforming my RNA-Seq count data using voom function of limma
package. I need to get the transformed data using the log normalized
data (voom.object$E) and the weight matrix (voom.object$weights). I am
interested in applying multivariate analysis such as classification
and clustering analysis, instead of differential expression. Thus, I
want to ask if there is a way to get this transformed data, to work
with this data outside limma package.
Thanks,
Gokmen
-- output of sessionInfo():
R version 3.0.3 (2014-03-06)
Platform: x86_64-apple-darwin10.8.0 (64-bit)
locale:
[1] C
attached base packages:
[1] parallel stats graphics grDevices utils datasets
methods base
other attached packages:
[1] edgeR_3.2.4 tweeDEseqCountData_1.0.9
BiocInstaller_1.10.4
[4] ReadqPCR_1.6.0 affy_1.38.1 limma_3.16.8
[7] Biobase_2.20.1 BiocGenerics_0.6.0
loaded via a namespace (and not attached):
[1] affyio_1.28.0 preprocessCore_1.22.0 tools_3.0.3
[4] zlibbioc_1.6.0
--
Sent via the guest posting facility at bioconductor.org.
You ask how to get the data from voom, but you clearly already know how to get voom.object$E and voom.object$weights, which is the output from voom.
Are you hoping that there is some way to combine these two quantities, so that the weights are no longer needed? No there isn't. The whole point of the voom method is to keep the log-cpm values and weights separate. As the published paper explains, the output should be used with software that is able to make use of the precision weights.
If you want to obtain log-cpm or log-rpkm values for clustering, then I suggest the cpm() or rpkm() functions in the edgeR package (using a largish value for the prior.count).