I am working on detection of Streptococcus pneumoniae in clinical samples like blood and CSF. I have been doing the assay based on the primers and probe recommended by CDC. However I reduced the reaction volume to 10uL instead of 25 uL, and increased DNA template to 2uL/10uL reaction volume instead of 2.5uL/25uL reaction volume. Initial run with a known positive had Cp value 23 and the NT (no template) had no amplification whatsoever.
After my second run, the NT samples started to give Cp 35.00 along with some of my samples. The article I used as a reference says Ct > 40 should be considered as negative. I have also attached the article for the reference.
Trying to find out a possible explanation, I searched in Researchgate, where some pointed out that Ct 31 - 32 should not be considered.
So I am unable to understand where to begin with. Kindly tell me how to troubleshoot this issue.
Thanks.
https://www.researchgate.net/post/Can_someone_advise_on_qPCR_woes_high_Ct_values_some_amplification_in_NRT_controls
You can determine this empirically: simply take a PCR product, gel extract it (so you know it's a pure prep), quantify it spectrophotometrically and thus calculate number of molecules.
Then use this in a dilution series with your qPCR primers, from say 10^9 molecules per well down to 10^0 or 10^-1 molecules per well (you can also use the resultant curve to determine the efficiency of your reaction)
Most well-designed primer pairs will happily work down to around 100 molecules per well, but become increasingly inaccurate at dilutions below this (stochastic partitioning effects).
A plot of log(molecules) against Cq should have a gradient of -3.32 (if 100% efficient) and a Y-intercept that corresponds (at a rough approximation) to the Cq of a single molecule. In practice, you'll see the datapoints becoming far too variable to be reliably quantified as you approach this Cq. You can quantify lower transcript levels (down in the 28-33 range), but you need to use additional replicates: rather than 3 replicate wells, use 5 or 6.
You are a messy lab worker. These new kits are extremely sensitive, the way PCR reactions are. Just popping the lid on your reagent tubes sends a burst of molecules into the air, and if you are not VERY careful, pretty soon your beard and shirt sleeves and lab pet hanging over your bench become sources of target DNA or whatever. Read the literature on the isolation of samples from prehistoric animals, including humans, to get an idea of how you must shield each reagent and each reaction tube.
Cq of 35 roughly corresponds to amplification of a single transcript: in practice, a Cq of 35 will be primer dimers/mispriming/stochastic noise. A melt curve should confirm this (or you could run your post-run samples on a gel).
Cq values in the 30-35 range typically correspond to transcripts numbers in the "1-20 molecules per well" range, which is considered too stochastic to be reliable: with only 10 molecules per ul, say, it's entirely possible to get unequal partitioning (5 molecules in one well, 14 in another, 8 in a third, for instance).
I usually trust Cq values as high as 28, but prefer my Cqs to be in the 18-26 range.
10ul reactions should be just as viable as 25ul reactions (I use 10ul routinely) but be careful about the amount of cDNA you're adding: cDNA synthesis reactions often contain a variety of chemicals that are deleterious to downstream reactions like qPCR, so if you add too much you'll impair your PCR.
For reference, I typically prepare 800ng of RNA in a 10ul cDNA synthesis and then dilute it ten fold. I use 1-2ul of this final dilution in a 10ul qPCR reaction (so 8-16ng cDNA per well, assuming 100% conversion).
You could try diluting your cDNA and running those (so neat cDNA, a 1/5 dilution, a 1/25 dilution, a 1/125 dilution or similar): this should identify any inhibitory effects.
Finally, you say you're using this diagnostically, so if some of your samples give you Cq values equivalent to 'no cDNA' reactions, the suggestion would seem to be 'these are negative samples', surely?
Dear John
I consider your answer very interesting, could you please add a reference supporting your claims. Or, do you think this should be proven experimentally?
You can determine this empirically: simply take a PCR product, gel extract it (so you know it's a pure prep), quantify it spectrophotometrically and thus calculate number of molecules.
Then use this in a dilution series with your qPCR primers, from say 10^9 molecules per well down to 10^0 or 10^-1 molecules per well (you can also use the resultant curve to determine the efficiency of your reaction)
Most well-designed primer pairs will happily work down to around 100 molecules per well, but become increasingly inaccurate at dilutions below this (stochastic partitioning effects).
A plot of log(molecules) against Cq should have a gradient of -3.32 (if 100% efficient) and a Y-intercept that corresponds (at a rough approximation) to the Cq of a single molecule. In practice, you'll see the datapoints becoming far too variable to be reliably quantified as you approach this Cq. You can quantify lower transcript levels (down in the 28-33 range), but you need to use additional replicates: rather than 3 replicate wells, use 5 or 6.
Dear Murty,
often we face NTC amplification after certain repeated use of primer and NFW working solutions. i suggest, try to make small aliquots of your working solutions to avoid NTC amplification. Determination of NTC amplification by Cq values depends on the qPCR machine what you are using. for examlpe, i am using ABI stepone plus qRT-PCR mechine, which suggest Cq values above 35 will not be considered as true amplification. u can practice 10 ul reaction volume for your amplifications.
gud luck
- Badges
- Science topic