# efficiency by std curve v/s sample-by-sample efficiency - for Corbett 3000 users (Sep/07/2008 )

Hi guys,

has anybody ever used the so called "comparative quantitation" function in the Corbett software?

I'm checking the amplification efficiency of my primers with the traditional serial diluitions and standard curve method and I got E= 1.61.. that's bad.

but reading the corbett 3000 manual and "Real time PCR" by M. Tevfik Dorak, I found out that the Corbett Software can "automatically determine the real-time efficiency sample-by-sample". I applied this function to the above samples above and I got an amplification=1.99 so the corresponding efficiency would be 0.99 or 99% (that's sooo gooood!). Why are the two efficiency values so different if they come from the same samples?

but overall, which value should I trust? I'd like the 2nd so much!!!!!!

:-)

Thanks a lot,

Raffaella

Hi,

as far as I know the efficiency determined by serial delution usually overestimates the efficiency (for reasons see also Tichopad et al. 2003). For validationg my reference genes I also use the second derivative method but I use three replicates per reference gene to have an average effeciency per gene and a control for each run. Both methods seem comprehensible to me, but I prefer the run to run effeciency determination just for having more control over what's happening.

Not much help, sorry...

Jan

has anybody ever used the so called "comparative quantitation" function in the Corbett software?

I'm checking the amplification efficiency of my primers with the traditional serial diluitions and standard curve method and I got E= 1.61.. that's bad.

but reading the corbett 3000 manual and "Real time PCR" by M. Tevfik Dorak, I found out that the Corbett Software can "automatically determine the real-time efficiency sample-by-sample". I applied this function to the above samples above and I got an amplification=1.99 so the corresponding efficiency would be 0.99 or 99% (that's sooo gooood!). Why are the two efficiency values so different if they come from the same samples?

but overall, which value should I trust? I'd like the 2nd so much!!!!!!

:-)

Thanks a lot,

Raffaella

Hi Jan,

thanks for your reply,

but with second derivative method do you mean the comparative quantitation (involving the take off points etc...)? If so, you are saying that I can trust the almost perfect efficiency value and discard the poor efficiency I got from the diluition series and relative standard curve?

I'm sorry for this dumb question, but it's to double check... I'm a bit confused about the way to go...

Thanks a lot,

Raffaella

Hi,

to be honest I am more confused about your bad efficiency values from your dilution series. In my case the second derivative method that is used in the "comparative quantitation" function always gave lower efficiencies than my calculations with the standard curve. And if I'm not wrong this is the way it should be.

I don't really know why it is the other way round in your case. Did you change something in the protocol that might influence the efficiency? (amount of template, amount of SYBR....?)

My suggestion would be repeat both experiments with exactly the same ractions and see what happens (If you haven't already done it...)

Sorry, I do not really have a clou, why your efficencies differ that much. It should be the other way round. All I can say that I trust and use the second derivative method, but in my case the differences are not that big (~ 0,15).

Good luck!

Jan

p.s. sometimes in real-time PCR it just helps if you try to wearing two different socks on your feet....

thanks for your reply,

but with second derivative method do you mean the comparative quantitation (involving the take off points etc...)? If so, you are saying that I can trust the almost perfect efficiency value and discard the poor efficiency I got from the diluition series and relative standard curve?

I'm sorry for this dumb question, but it's to double check... I'm a bit confused about the way to go...

Thanks a lot,

Raffaella

Sorry, but I'm about to try and go to bed so no references.

The SDM derived efficiency is meant to be too low (taken at the end of the exponential period). The standard curve measure of efficiency is meant to be too high. The window of linearity method (google linreg) is meant to be very accurate, and the latest linear regression of efficiency method best of all (if your curves fit the model). Also, it's best to use the average of all the linreg efficiency data for a particular run, rather than adjust on a reaction by reaction basis (paper titled 'evaluation of different methods for estimating qpcr efficiency' or something like that has a comparison.

I use the linreg method to get the average efficiency of the reactions (and to exclude reactions with outlier efficiencies), and use this with the Pfaffl method for my analysis.

I'm even more confused than you Jan because the two different efficiency values are for the same reaction. I mean, afer the run with the diluition series I drew the std curve and I got E=1.61 but the reading the manual I found the comparative quantitation method and I used it on the same run data and E was 1....

at the moment I sent my.rex file to corbett and I'm stiil waiting for their comments...

from what I've read the SDM derived efficiency is taken before the end of the exponential period as it is calculated based on the take off cycle and the next 5 ones, but I'll have a look to the linreg method yoe were quoting maset... or maybe I'll throw my samples out of the window!!!

I'm keen too wear even two differnt shoes to get this work done!!!