DE analysis between different species

[super0925]

Hi All
I have two group of animals. cow and sheep
GroupA: cow1 cow2 cow3
GroupB: sheep1 sheep2 sheep3
All of them have finished the Tophat mapping.
Could I do the DE comparison between two groups?

(1) So far, I don't think Cuffdiff could do this, am I right?
(2) For edgeR/DESeq2,could I firstly get the counts and then uniform the Ensembl ID between cow and sheep (transfer them from cow to sheep or vice versa) , and then do edgeR/DESeq2. Is this OK?

[dpryan]

Your steps will be:

(1) Define a set of IDs that will be constant between the two species. This will allow you to derive counts for each.

(2) Define a set of constantly expressed genes that can be used for normalization. Without this, you have no way to compensate for things like a species-dependent alignment bias. You may also need to account for transcript length differences between the species.