Colleagues,
Do you know if selection of reference genes for qPCR needs to favour those reference genes that have expression levels (Ct values) in the same range as the target genes? If so what are the limits of this 'same range'?
What is the mathematical implication of using relatively very highly abundant reference genes to normalise low copy number target genes, provided we are happy with the stability of the reference genes; and that the coefficient of variation between the two remain comparable? Could there be another issue to consider?
Has anyone ever studied this empirically? Any references please?
I have seen at least one publication (and one blog) recommending dilution of the cDNA sample to 'lift' the Ct values of the reference gene towards that of the targets. This can result in grave error, but would it be justifiable at all?
Thanks a heap!
Hi
To pick up a reference gene (house keeping genes) you have to run several ones on your samples to find out which of them is constant and gives you the lowest Ct value. At present the majority of companies that supply qPCR products provides a set of housekeeping genes that you can test easily on you samples to pickup the most suitable for you. Please check the provided links.
Regards,
Bassem
http://link.springer.com/article/10.1007%2Fs13353-013-0173-x
http://www.cell.com/trends/genetics/pdf/S0168-9525(13)00089-9.pdf
http://www.tau.ac.il/~elieis/HKG/
http://www.sabiosciences.com/rt_pcr_product/HTML/PARN-000A.html
http://www.sabiosciences.com/rt_pcr_product/HTML/PAMM-000A.html
http://www.sabiosciences.com/rt_pcr_product/HTML/PAHS-000A.html
The purpose of a reference gene is to ensure that your samples have been loaded in relatively equal amounts (cDNA amount). So, it needs to be abundant so you have a relatively low CT value.
Also, bear in mind that if you dilute to "lift" the CT value you should do the same for your gene of interest and if that gene has a low copy number you might have problem getting accurate results.
Finally, make sure your experimental conditions do not effect your reference gene - that's why its best to run a panel to figure out what works best for your experimental conditions.
Thanks Lilian and Bassem,
I agree with your answers above. Thank you very much, for your time too. Now suppose we have run the checks to find which of the reference genes are left with a small bunch that can be combined to give stable Internal references, something remains, should we now go down to the expression levels? Should we prefer one or ther just because of their expression level?
Aman
I don't think you should combine reference genes , it can easily lead to questionable data. Instead of combining the reference genes, choose one and increase the concentration of the template for the qrt-pcr.
How would you increase template concentration? Adding more sample? For what reason?
OK, let us suppose we had candidate reference genes that have different expression strengths, and you have to choose one. Which one would be preferred, one that has the same expression strength as the genes of interest or those that are stronger or weaker. Why?
You need to run at least two reference genes for every qPCR. Read this reference for some important tips: http://www.plantcell.org/content/early/2014/10/31/tpc.114.130641.full.pdf+html
Colleagues,
Thank you for your insightful answers and relevant references. I managed to download most.
Yes for the definitive experiment we consider at least two reference genes but then compare combinations of the most stable candidates.
Again, thanks for your time and views.
Most appreciated
- Badges
- Science topic