Comparing FPKM across samples

[jk1124]

I have two RNA-Seq samples, one which I performed using Spike-In Controls and the other without Spike-Ins.

I mapped and assembled (Tophat, Cufflinks) the first using a reference assembly (-G option, no novel) and bowtie index with spike-ins concatenated to the end. The other I mapped and assembled using a reference assembly (-G option again, no novel) and bowtie index without spike-ins concatenated to the end.

Can I compare the FPKMs I get from each assembly?

Or I could perhaps reassemble my first sample using the reference assembly without Spike-Ins, and provide the mask-file option with the Spike-In transcriptome annotation, so that the Spike-Ins get discarded. Would that work?

I would appreciate any insight. I am new at this.

[jk1124]

Alternatively, might be a better idea to use a count based method instead?