RNA-Seq compare edgeR, cufflinks, Partek

[schaffer]

Hi,
I have compared RNA-Seq in 3 different platforms and get 3 different lists of differential gene expression. Has anyone done this kind of comparison?
What do you think about this?
Lana

[severin]

how different?

Are the 3 lists from the same set of data?
What percentage of genes are the same? different?
Are the programs using different statistical tests to make the comparisons?
Do they all incorporate/deal with replicates the same way?
What are the cutoff p-value used in each? Do they all correct for FDR?

I would be surprised if you had no genes in common between the three programs.

[mgogol]

Are there really NO genes in common? Do they share gene ontology terms or pathways?

Unfortunately, I think it's pretty common in bioinformatics for the analysis techniques to have large effects on the final list of genes and their rank. You could test genes from each list with qpcr to try to see which are the most solid.

[Simon Anders]

There are a couple of threads that may be of interest, including these:

http://seqanswers.com/forums/showthread.php?t=12639

http://seqanswers.com/forums/showthread.php?t=10797