Validation of microarray data by qPCR - (Jan/11/2012 )

Hi! I have to validate my microarray data performed in Affymetrix chips by qPCR (SYB green) and I have some questions about selecting genes. I find the literature a bit confusing regarding this aspect. How many genes do I need to validate? The majority of interesting genes for my research are over-expressed. So, do I also need to perform qPCR for down-expressed genes? Is there any biological or technical difference regarding validation between up- and down-regulated genes? I’ve found in literature that array and qPCR have lower correlation for down-expressed genes because qPCR loose efficiency at later cycles (Ct), where genes with low expression levels respond (Morey et al., 2006­_ Biol. Proced. Online 2006; 8(1): 175-193). Is this correct?

I would really appreciate some help.

Thanks

You should validate all key genes. qPCR is generaly took as more reliable, it's true that problem may be with less abundant genes, but those in the safe range (Ct 15-30) should be better than microarray.
Microarrays acts as some sort of guide. If you see a pathway that is overexpressed, select the usual gene of them and validate. If there are genes especially interesting, validate them. Depends what you want to formulate out of your results. If you goo deeper in some pathways, then validate all genes from there. I don't think it depends if gene is up or down regulated, microarray doesn't have the same weight as good qPCR validated results.