[R] Differential Co-Expression ?

Question

[ali]

Hi everyone , before doing the Differential Co-Expression between Tumor and Healthy I have to export the count matrix from a DGEList , taking into consideration the TMM normalization on it. Should I use :

counts <- cpm(y) 

or :

counts <- cpm(y,prior.count=1,log = TRUE)

?

[ATpoint]

That depends what the package you use instructs you to do. Do you need log-scale or not?

[ali]

I use edgeR , that is the point, I dont know if I need log-scale or not.

[ATpoint]

If you are doing the normal edgeR differential expression workflow then you should do what the manual instructs you to do.

https://www.bioconductor.org/packages/devel/bioc/vignettes/edgeR/inst/doc/edgeRUsersGuide.pdf

In short: You never have to provide normalized counts explicitely. Run calcNormFactors and the functions will fetch the counts and normalization factors internally. Please follow the manual precisely.

[ali]

But I have a DGEList where I applied :

y <- DGEList(counts = counts, group = group)
y <- calcNormFactors(y,method = "TMM")

So in order to extract the matrix i did use that function. The DGE is done and finished , now I need to do a Correlation analysis so I need the count matrix normalized before that is why. I followed all the steps but that was for DGE. Now I need to do Network Analysis and I just wanted to obtain the matrix from the DGEList I used. This is the complete name of the analysis : Differential Co Expression Analysis of Gene Expression.