some qPCR questions (naive ones included)... - first pitfals etc... (Nov/17/2005 )

hi, i am using the takara sybrgreen extaq kit on a mj opticon2 cycler.
i have been trying to set up my first qPCR experiments and stumbled upon some questions/problems, i hope someone can help me out.

1. i am using the PFAFFL method for the determination of relative gene expression.
when i want to determine the ratio between two samples, in one of my samples
the curve will never come up. that means i will never be able to determine a factor, or should i
just set the ct value to 50 ??? however this seems strange to me, because i have a faint band
when i am doing conventional RT PCR.

2. why do people say that the threshold should be set equal for the target and the reference
gene? i don't quite get the rationale behind that, it shouldn't make a difference as long as the treshold lies in the linear phase of amplification, right?

3. i have experienced large interassay variations in ct values. what can be the main reason for this?

4. when i run my pcr products on the gel, i often experience unspecific amplification in one of the triplicates. but what can be the main reason for this? i pipetted it virtually simultaneously. could it be the primers, or the template itself (maybe partly degraded)???

5. i have experienced additional peaks in the melting curves in both cases, when the template is too concentrated, or when the template is too dilute, however this never happens with the hprt1 primers i am using, which eather give a nice or no signal.

6. will i have to make a new standard curve for efficiency calculation when i am increasing the annealing step temp of the reaction, to make the signal more specific?

7. how can the efficiency of a pcr reaction be more than 2??? it can't do more than double right?
maybe someone has a nice explaination for this.

8. what is the best way to store your diluted cDNA ? i used dh20 and 4 degrees, what might not be the best.


a lot of question, sorry, but maybe someone will have the time to give me a few answers.

thanks a lot in advance.

best wishes

martin

---

hi doesn't really have anyone an answer to at least one of the questions?
...can't be possible, right?

i would be really thankful for some answers!
cheers,
martin

---

well
i'm not dealing regulary with QPCR so i'll try to correctly answer to questions...

QPCR is based on comparison between two conditions (reference and target). So it's a tough point to set up a good comparative assay.



did these assays were performed same day? (i mean two assays with the same base sample or did you do these assays in different days?)

when you say virtually simultaneously do you mean that you made a mix divided in three? or did you pipet three times? in last case a change in glycerol final conc. (during pipetting of the enzyme) can affect general specificity. But it's quite surprising...


i think so

if i understand this point good, you mean that after one cycle, one mlecule may drive to more than 2 molecules??? As i know it's theorically impossible. And usually it's close to 1.8


i didn't store my cDNA. But if i had to, i would say that -20 is better than 4°.

hope that clarifies for a little.
fred

---

how can the efficiency of a pcr reaction be more than 2??? it can't do more than double right?
maybe someone has a nice explaination for this.

sounds like you've got a problem with your pipetting technique. efficiency schould be ~1. this is based on the standard curve. my guess is that you've imput too much cDNA in the standards comapred to the amount you typed in for each sample. ie for the 1/10 sample, you've actually put in 1/6. this would change the slope of your standard curve, which is what the efficiency is based on.

i have experienced large interassay variations in ct values. what can be the main reason for this?

things change. the amont of primers might be slightly different, the amound of cDNA might be slightly different. tubes might be a little dusty, or oily, or who knows what. try to keep things exactly the same through each run. this should limit the variation.

what is the best way to store your diluted cDNA ? i used dh20 and 4 degrees, what might not be the best.
-20`C. put it in aliquots, and freeze. it'll go off otherwise.

will i have to make a new standard curve for efficiency calculation when i am increasing the annealing step temp of the reaction, to make the signal more specific?

yes.

fred was right, just reitterating it all for ya.

---

I am fairly new at qPCR, but I have the same problem, with excessively high efficiencies. I just watched a webcast ("How Reliable is Your qPCR Data?" presented by Drug Discovery and Development) in which that exact question was addressed.

The experts on the panel said that high efficiencies happen when the amount of template in your reaction is very low, resulting in high Ct's - so that at your high dilutions, the Ct's get "scrunched". They called this problem "saturation", but I don't understand exactly what they mean by that...does anyone else?

Anyway, they said that you need to throw out the "scrunched" Ct values at the end and only use the lower values to generate your efficiency curve.

I hope this is helpful!

(I tried to post the link to the on-demand webcast, but it would not show the whole link in the preview...! Sorry.)

---

Sorry I don't get the question "curve will never come up" ???

It should not be equal. Threshold as well as baseline should be same for a given gene in all groups compared, but not between different genes.

It cannot be. If you get it higher then 2 that means something is wrong with assay. Too much or too little template (optimize nucleic acid extraction and RT), Pipeting error (calibrate, lerger volumes), also they recommend to omit inhibited dilusions in validation test, in other word remove outliers.


I store it at -20, minimize thawing and freezing.

---

hi thanks for your replies

with the curve will never come up, i mean that there is ???soo little template?? that the curve will never cross the treshold line. if that happens, you set the ct to 50 or higher (i.e. such a high treshold would indicate that there is nothing in there....) for calculations...


i don"t think that there is something wrong with the assay, if the efficiency gets higher than 2.
check this paper for reference:
pfaffl paper

greetings

martin

---

Pfaffl paper link doesn't work. Maybe if elongation is too long or too much primers or template polymerase can copy more then once. Any way assay needs optimisation.

If there is no proper curve then any quantification would be impossible. Even when Ct values are above 35 quantification gets very dificult, becouse of efficiency problems.

---

Here is the Pfaffl paper..

---