Different qPCR Efficiencies (slopes) Between Standards and Experiemental

17 posts / 0 new
Last post
mayrberg
mayrberg's picture
Different qPCR Efficiencies (slopes) Between Standards and Experiemental

I'm running SYBR green qPCR of Flavobacterium 16s rRNA gene copy number from genomic extractions of soil samples.

I have a genomic standard template that I include in each plate for an in-plate standard curve.

The problem I'm having is that the slope of the standard curve does not match any of the serially diluted samples. 

I'm guessing that is why my calculated copy numbers for the experimental samples are not the same, even though it's a serial dilution of the same template DNA.

I'm hoping someone can point me to a good text regarding how I can mathematically account for the differences in replication efficiencies and still get a meaningful information from my qPCR.

Ivan Delgado
Ivan Delgado's picture
 

 
Hi mayrberg,

I am not sure if I understand what you mean when you say that the "slope of the standard curve does not match any of the serially diluted samples". Please go ahead and post some Ct values for your standard curve dilutions and experimental samples to get a better idea of what is going on. Any additional information, like how much template you are using, would also be useful.

Cheers,

Ivan

mayrberg
mayrberg's picture
For example...

For example...

I have a serial dilution of my standard which was replicated with an efficiency of 89% (m=-3.61).

I have a handful of experimental samples that I serially diluted.  The replication efficiency of those samples were between 88% (good; m=-3.65) and 110% (not good; m=-3.12).

If my samples are replicating with a greater efficiency than the standard, then I can not use my standard curve to directly back calculate gene copy number because the Ct is going to occur sooner for the samples than it does for the standard.  (Different efficiencies.)

I'm sure it's just some simple conversions to adjust for differences in efficiencies....

Ivan Delgado
Ivan Delgado's picture
 

 
My first concern with the results you are getting for your standard curve is that an efficiency of 89% is not good. While it is normally acceptable to have efficiencies between 90% and 110%, the closer you are to 100% the better. I personally have never used a qPCR assay with an efficiency lower than 98%.

Assuming you want to go forward with assays that have efficiencies lower than 90%, which I do not recommend, the efficiency you are getting for your standard curve samples and experimental samples should be similar. In my book that means less than 2% different (for example, 97% and 99% efficient). Getting a range of 88 to 110% efficiency is definitely a problem.

I am going to assume that the way you prepared your standard curve samples and your experimental samples is different. If that is the case, that may well be the reason why your slopes vary this much. You need to prepare your standards and experimental samples in the same way. Ultimately your RNA, and cDNA, need to be of very high quality otherwise you will have problems with your qPCR.

Alternatively I recommend you look into your assay too. As I mentioned above, an assay with an efficiency that is more than 10% removed from 100% is not good and will very likely cause you problems in the future.

Finally, your low efficiencies could also be due to using too little or too much template. If you look at the Ct values of each one of your dilutions, they should be around 3.32 Ct points different from each other. If they highest or lowest dilutions are significantly different from the middle dilutions, then you are likely having a problem of too much or too little template.

Diana J
Diana J's picture
Hello Ivan!

Hello Ivan!
I've been reading your posts and you seem to be very well informed about qpcr. I'm struggling with my standard curve as well  and you might be able to help me finding out what's going wrong.

Here is the situation.
Purified viral DNA standards, 25000   -  2500  -  250 -  25 and 2.5 picograms in duplicate
Efficiency, between 64% and 74% (consistently)
Cts 17.2   -  21.3  - 25.5  - 29.7 - 33.9
R2 usually 0.99-1
Slope -4.48

I don't know why of these results, according to the R square it shouldn't the pippeting but other operators have got better results (not always).

Thank you very much for your help.

Diana

Ivan Delgado
Ivan Delgado's picture
 

 
Hi Diana,

First of all, I am sorry to hear that you are having problems with your assay. One of the tricky things about molecular biology is that sometimes things simply work right away, and some times they give you problems even when you do everything correctly. For instance your dilutions series is ideal.

Here are my comments about your results: 

1. Your pipetting technique seems to be good, although I would be more comfortable if you provided R2 values with 3 decimal points. An R2 of 0.990 is acceptable, but not very good, and very different from an R2 of 0.998, which would be for all intended purposes good enough for any application. 

2. How did you determine the concentration of your virus DNA sample? In my experience viral DNA can be a tricky sample to quantify and I just want to make sure your calculations are solid.

3. How did you prepare your different DNA dilutions? The fact that you are getting a slope of -4.48 is odd. What this is telling you is that the difference between each 10-fold dilutions is much lower than the ideal 3.32 PCR cycles. In other words, the efficiency of your assay is very low as you pointed out. Low efficiency typically occurs when the assay has not been optimized. Did you optimize your assay? Alternatively low efficiency is a result of primer dimers. Did you check for the presence of primers dimers by running a melting curve analysis using SYBR Green?

Provide me with some more information about your assay and maybe I can be of additional help. I would like to see things like: a. how clean is your amplicon in a melt curve analysis; b. optimization of the assay to determine at what concentrations of primers you get the best signal; c. any other alternative assays you testing to determine which assay worked best for you (I always test at least three different assays and chose the best performing one); d. How good is the quality of your viral DNA and how was it quantified and diluted.

Cheers,

Ivan

Diana J
Diana J's picture
Ivan: thank you so much for

Ivan: thank you so much for your interest and prompt answer.
I just arrived to the research team and it took me a while to gather the answers for your inquiries.
Well, here it goes.
1. The R2 decimals, the lowest I've got was 0.993, but with R2=1 the same low efficiency occurred

2. The DNA concentration was determined by the company that purified the viral DNA.  So, I really don't know

3. Dilutions preparation, I dilute my first standard in order to get a 10 ng/ ML concentration (it comes 27.1/ML) and then I take 5ML of this dilution and put it in 45 ML of water , I do the same with this second dilution, take 5 ML and put in 45 ML, and so on.  Always vortexing and spinning between dilutions.
Optimization, well, it is not formally optimized, because it ran pretty accurately at the beginning with the suggested protocol.
Melting curve, beautiful, no extra picks, very tidy.

Anything in mind?

Thank you again, sincerely

Diana

Ivan Delgado
Ivan Delgado's picture
 

 
Hmm, everything you describe sounds good. You pretty much can trust commercial DNA. Single peak in a melt curve analysis. 10-fold dilutions.

Do you have any other assays, and associated standard DNA, that give you >90% efficiency when you run a standard curve? This is what I am thinking: this may be an assay-specific thing, so no matter what you do the primers you are using will never give you the results you need. In such situations I would feel much more comfortable running another assay I know works to make sure all my reagents, and instrument, are working as they should. If that is working, then you can be pretty sure your assay is not good. 

Ivan

Diana J
Diana J's picture
thank you

thank you

sama
sama's picture
 I am facing a problem with

 I am facing a problem with my qpcr. I am optimizing 90 genes with qpcr doing standard curve. I am doing it in duplicate. I got many of the results with efficiency higher than 110, sometimes 147 or 150 or higher. I don't know what to do. Given that I optimized my primer previously with regular pcr
Please help,

Ivan Delgado
Ivan Delgado's picture
Hi Sama,

Hi Sama,

My first question to you is: did you design these assays for qPCR or for PCR? If you did the later, then most likely you need to redesign your assays since qPCR assays are not the same as PCR assays. 

If you did design your assays for qPCR then please share more details about what you are doing to optimize these assays.

sama
sama's picture
Thanks for your reply.

Thanks for your reply.
I designed my primers using Primer3, and then I optimized them with regular pcr because it has cheaper reagents to make sure that I got one band and to know what is the optimal anealing temprature. I spent about 4 months optimizing them . Now I moved to do quantification for my genes using the qPCR, most of the standard curves efficiencies that I got are above 110 to 150. I used 10 fold dilutions, but at the 1:1000 some of the genes don't amplify at all. So I tried to do 1:1, 1:5 etc (5 fold diutions), but still the efficiency is 150.

I am using Syber green, my primers concentration either 5pmol/ul or 6.25/ul I am using 1ul of each primer in 10 ul reaction. Are the concentrations good?

I don't know what to start with to trouble.

Thanks

Ivan Delgado
Ivan Delgado's picture
I understand your situation,

I understand your situation, but unfortunately standard PCR and qPCR are very different technologies. Any optimization you did using standard PCR will likely not work for qPCR. To start, qPCR requires amplicons that are no more than 100 to 200 bp long.

I am sorry, but it seems to me that you need to start from scratch and re-design your assay specifically to work for qPCR. 

If I am missing something do not hesitate to let me know. 

Good luck

sama
sama's picture
 Yeh you are right about bp

 Yeh you are right about bp lenght, I concederd that when I designed my primers. All my segments were less than 250 bp.

Ivan Delgado
Ivan Delgado's picture
Unfortunately then I am not

Unfortunately then I am not sure why your assays are not working. It is possible that the problem comes from your DNA too. If at all possible get DNA that is of very high quality, even buying commercial DNA if need be.

I do not see why assays designed using Primer3 would be an issue, but it would also not hurt to design an assay or two on software that was written to design qPCR primers. Unfortunately I do not have experience using Primer3 for qPCR assays. 

sama
sama's picture
 Do I need to add the primers

 Do I need to add the primers to my master mix when I do repicates? This might cause a problem adding them seperately.

Ivan Delgado
Ivan Delgado's picture
I would prepare a single

I would prepare a single master mix that contains all components, including primers, and use that same mix on all qPCR reactions.

To answer your question: yes, adding your primers separately may be causing problems.