We are sequencing through genewiz and see that they offer PhiX spike in controls. For normal RNA-seq (without major concern of limited sequence diversity), are Phix spike-in's necessary?
When sequencing, a PhiX spike is a very good way of showing that the sequencer is working properly (regardless of the diversity of your sample). If the PhiX is present during sequencing, but your sample is not, then the library construction of your sample has not worked. For this reason I would imagine that they will put a small (maybe 1% spike) in all samples that they run.
If your sample is low diversity/has many repetitive regions etc the higher the spike should be. Would your sequencing run be a pool of many libraries/samples?