RNA-SEQ, Gene duplication, and gauging system regulation

[Biomicrogen]

Hi All!

So I have a few questions here:

I am comparing gene expression between two lifestyles of a novel organism, which I have a draft annotated genome for. There appears to be several copies of many different genes distributed at various loci on the genome, this could be an assembly artifact but I'm not inclined to believe that as the genomic context around most of the duplicated genes is different.

1) If I have several annotated "copies" of a gene which my transcripts map to, and say there is variation in the Log2 fold levels of each (some up some down, all significant) of the "copies", can I add these numbers to get an overall expression change of this one gene if I think it is a duplicate? However, perhaps this is a biologically relevant change than by doing this I would loose valuable information.

2) Along these same lines, if I am looking at a system, say CDS involved in photosystem I, would it be biologically relevant to add all the fold changes of all genes in this system and say that Photosytem I is generally X fold ip/down regulated?? It seems a fairly straightforward assumption but perhaps I am missing somethiing?

[dpryan]

1) Sure, I've used the same trick with other transcripts that happen to appear multiple times in the genome.
2) No, that's not legit. What you're looking for there is called "pathway analysis", though it won't give you a fold enrichment (if you needed that, you might just report the median change, though that is itself not without issues).