Cufflinks followed by Microarray-like statistical test ??

[taozuo]

For cufflinks, I know that cuffdiff can perform statistical test for identifying differential expressed genes. And I also know that there are several other tools could conduct statistical analysis for Rna-seq data, like DEseq.

But, is it reasonable to do Rna-seq like this:

1. Tophat to align the data
2. Cufflinks to get normalized expression values (FPKM).
3. combine the FPKM values from different samples
4. perform statistical test use the method implemented in Genechips (Affy).

If not, what are the disadvantages to make it impossible? Thanks.

[Dario1984]

Take a look at the current version of limma. It does what you need.

Changes in version 3.10.0:

New function voom() allows RNA-Seq experiments to be analysed using
the standard limma pipeline. An RNA-Seq case study is added to
User's Guide.