Calculation issue for inconsistent results in triplicates in Taqman rPCR. - (Oct/21/2008 )
Hi All,
This topic is highly relevant to a topic discussed in the forum recently. My ongoing project is to detect very low copy of target genes from urine. The Ct are around 40. I can not increase the template amount into PCR reaction mix simply the total DNA is very limited. I split one sample into 3 to form the triplicates and believe that my pipetting skills are good enough.
The problem encountered now is inconsistent results in triplicates from those samples with extremely low gene copies. One well of the triplicate may show a Ct of 38, another 39, while the third one of undetectal. The worse is seen in those samples with one of triplicate with a Ct of 38, and the other 2 wells of undetectable. I'll be very much grateful if anyone could suggest how to quantify those samples, if we are not ready to sacrifice these quantitative data into qualitative data, i.e., positive vs. negative.
Help me and thanks in advance!
This topic is highly relevant to a topic discussed in the forum recently. My ongoing project is to detect very low copy of target genes from urine. The Ct are around 40. I can not increase the template amount into PCR reaction mix simply the total DNA is very limited. I split one sample into 3 to form the triplicates and believe that my pipetting skills are good enough.
The problem encountered now is inconsistent results in triplicates from those samples with extremely low gene copies. One well of the triplicate may show a Ct of 38, another 39, while the third one of undetectal. The worse is seen in those samples with one of triplicate with a Ct of 38, and the other 2 wells of undetectable. I'll be very much grateful if anyone could suggest how to quantify those samples, if we are not ready to sacrifice these quantitative data into qualitative data, i.e., positive vs. negative.
Help me and thanks in advance!
You might be down to so few copies of your target that quantification would need very large numbers to be run. In each tube you would have different numbers, and many tubes would have none. I don't think any exact quantification would be feasible.
When you don't detect the target, is it that it wasn't present in the sample, or is it that something else amplified instead.
If you are using probes, you could try adding SYBR Green as well - if something amplifies starting at cycle 32 or something then maybe this is preventing you from detecting your wanted target.
You could add SYBR Green to your negative pcrs and then run a melt curve to see if anything has amplified. If nothing has amplified at all then clearly its a true negative and you didn't have any target in your input and you'll have to consider poisson distributions unfortunately
Cheers
TK
You could try a version of a limiting dilution assay?
Cheers
TK
Thanks, TK. We'll take your advice and use SYBR Green to exclude the possibility of competitive fragments.
rPCR7
Hi, Maset,
Thank you for the suggestion. Would you please explain a bit more on setting out of the experiments for the limited dilution assay?
rPCR7.
Hi, Maset,
Thank you for the suggestion. Would you please explain a bit more on setting out of the experiments for the limited dilution assay?
rPCR7.
Unfortunately I only frequent these forums late at night. Basically a limiting(limited) dilution assay dilutes an assay down to a binary state (on or off) and uses the ratio of negatives to calculate the actual calculation. For example in any given large dilution there could be 100 to 0 copies, but the distribution is mathematically governed.
It's primarily a microbiological protocol, so start looking there.
Taqman qPCR even can be used to detect the one/two copy DNA difference. Your primers and taqman probe may need to be redesign carefully so as to detect one copy of DNA molecule.
This is true in theory, but actually many other factors arise when working with very dilluted samples. For example, there are theoretically two copies in your sample, and you take half of it, so mathematics tell you, that you have one copy in reaction, so you're gonna detect it. But in reality there is big chance that you take both copies, or none and that's just the starting point, not even polymerase and other stuff works mathematically correct.
I've seen a limiting dilution assay somewhere, maybe some bulletins or something like that, but I can't find it now, there was a neat graph too.. but generaly said, you make a series of dillutions in big number of replicates (like 20) and count how many times the replicate comes positive at all, the number should be lower with each dilution. You put it into a graph and find a mathematical limit of it, that is the limiting concentration.
[/quote]
I've seen a limiting dilution assay somewhere, maybe some bulletins or something like that, but I can't find it now, there was a neat graph too.. but generaly said, you make a series of dillutions in big number of replicates (like 20) and count how many times the replicate comes positive at all, the number should be lower with each dilution. You put it into a graph and find a mathematical limit of it, that is the limiting concentration.
[/quote]
Thank you, Trof,
it sounds like a digital PCR method. Ok, I'll try to search for the protocol for it. Guess this method is good for research only. If it is applied to a clinical test, the test fees might be another issue since each sample will consume a lot of reagents.
rPCR.
Actually I think the main use of this is to estimate sensitivity of an assay, so it should be performed once before putting a new diagnostic assay to use. You need to know how sensitive your reaction is, especially in diagnostics. Of course in routine testing you wouldn't set up 20 reactions, but you can tell the sensitivity treshold.