Cross platform gene expression comparison and threshold choice

Question

[elb]

Hi guys, I have a question about how to choose a reasonable threshold to define genes to be expressed (not differentially expressed but simply expressed) to compare gene expression measurements from different platforms. In other words I have to check if my list of genes is expressed using an RNA seq dataset, using a data set of one-color microarrays and another of two colours microarrays. If for RNA Seq is generally "accepted" (but I know that it depends on the experimental condition) an rpkm >1 I don't know how to set a threshold for microarrays. Any suggestion about this?

Thank you in advance

Answer 1

[Devon Ryan]

There is no generally accepted threshold in RNA-seq. You can come up with a threshold using zFPKMs, but even that is a bit hand-wavy. For microarrays, you generally check either that the PM signal is sufficiently above the MM signal and/or that a probe is sufficiently above that of the background control probes. Not that these are sort of apples and oranges measurements, in that they're really not comparable.