isodiff vs genediff

[Cody Smith]

I have a basic question about the differential expression analyses. When is it appropriate to use either the list of differentially expressed isoforms or the list of differentially expressed genes?

I ask because in one of my experimental groups, there are about 1400 differentially expressed isoforms but only 5 differentially expressed genes. I understand that the analysis considers read ambiguity among gene isoforms, but it was still surprising to lose so many differentially expressed isoforms at the gene level. The list of differentially expressed isforms makes sense biologically when plugged into some higher order pathway analyses, so maybe I should just use the list of differentially expressed isoforms?

Any help is greatly appreciated.

Thanks