Protocol Online logo
Top : Forum Archives: : Real-Time PCR

Setting up and analysing real time PCR - (Feb/05/2008 )

Hi.

I've inherited a stragene MX3000-P machine and I'm teaching myself real time PCR.
I've read the ABIprism bulletin 2 http://hcgs.unh.edu/protocol/realtime/UserBulletin2.pdf on how to set up relatively quantify expression using seperate tubes (p27-29). Why on p28 is the cDNA of interest is diluted with yeast RNA to construct standard curves?

Also I'm using syber green, how should I set my standard curve up? I'd be very grateful for as basic a explanation as possible so I can replicate it accurately.

I should perhaps say, my experiment is looking for a change in a single gene before and after hormone treatment in the same cell type.

Many thanks

Sean

-seangbrown-

QUOTE (seangbrown @ Feb 5 2008, 07:05 PM)
Why on p28 is the cDNA of interest is diluted with yeast RNA to construct standard curves?


I guess primary to make you buy yeast RNA from Ambion wink.gif
But seriously, sometimes your cDNA of interest is mixed with a normalised amount of a "dummy" nucleic acid to ensure the same basic NA concentration (while the concentration of your template and the genes you look for will level down). This is to make a same conditions as you would have with a real low-expressed sample.

But I don't do it though, not on a relative quantification (if you're not determining detection limit). I usualy don't do dilutions at all, when I take amplification efficiencies of both genes as similar and use delta-delta Ct.
But if I want one, I just dilute the cDNA containing highest amount of target gene like 10 fold, several times (at minimum 3, 4 is better) and run them in triplicates. To count the amplification efficiency (that's the only thing you need standard curve for in RelQ) assign the dilutions numbers 1 (non-diluted), 0.1 (10 fold dilution), 0.01 (10 fold dilution of the 10 fold dilution) etc. as a copy number/concentration, and let the machine compute the slope and stuff.

PS: keep in mind, that these bulletins are made to make you spend more money on their reagents, so that they overestimate number of replicates (3 is enough, ABI sometimes recomends 5), reaction volume (I think 50ul is unnecessary for real-time if you can run in 20ul - check your machine manual for suitable volumes) and reagent losses due to pippeting errors (in my personal experience, 15% extra works fine). It's OK to go by the book first, but when you get it in hand, you can find ways how to economize.

-Trof-

I agree with Trof.

The only thing I would add, though, is that in order to be able to compare a house keeping gene to your gene of interest is to make sure that the PCR efficiency is the same. And to do that, you have to make a standard curve with both genes and compare the slope of the resulting curves. ABI's manual usually gives out good advices to do so.

-Madrius-