Opinion: What fold change is actually considered meaningful - (Mar/26/2013 )

Thanks in advance for replies.

We've been doing some qRTPCR in my lab and using ddCT to calculate relative abundance of a couple genes of interest on some cultured cells. We routinely see between 2-4 fold induction under one condition vs the control condition. It was statistically significant and repeatable, but my boss asked "whether or not it's actually meaningful."

Her point was that with RTCPR, if your replicate samples and then replicate wells for each sample are all close enough, a 1% raise in gene expression could be statistically significant, but it likely would not be meaningful. Thus, at what fold + or - vs a control can you consider meaningful? My boss suggested using a gene that we are confident gets upregulated and expressed under our experimental condition vs. control, and use that fold induction to compare to as "meaningful". But I don't know if this is helpful either, since different proteins/enzymes have completely different affinities/activities/whatever, and thus a different gene's expression would be completely arbitrary here.

Thoughts from anyone?

Also, to anyone who might suggest looking at protein expression from the cell culture for verification of expression...the reagents for what we're trying to detect are notoriously bad.

For microarray work a 2 fold change is considered relevant.

Apart from fold change (needs to be statistically significant) in mRNA level, you may need to consider other things determine whether the change is meaningful. For example, whether the change in mRNA translate to an increase in protein level, whether there are expected phenotypic changes associated with the function of the gene?