Ask me anything on dPCR? 

你問我答第二集

11. What’s the meaning of the confidence interval (CI)?

The confidence interval (CI) is a statistical value that is calculated based on the Poisson distribution. This value is mainly influenced by the number of molecules per partition. The CI is considered very good when we are in the suggested range of 0.5 and 3 molecules per partition and still pretty good around 0.05 to 5. But the value gets worse when found outside of this input range. The CI is mainly useful as an indicator of which dilution of a sample is in the suitable range for dPCR. If the CI is 10% or lower, then no adjustments are necessary. If the CI is higher, it would be better to use a bit more or a bit less DNA, if possible.

If you don’t adjust the CI, you cannot reach the most precise measurement. Does this mean your results are no longer usable? No. It means that your results are slightly less reliable with 10% precision at 0.1 to 20% at 0.001 and 60% at 0.0001. In other words, it becomes harder to measure small differences between your samples at the lower copy end. Bear in mind, that even with a lower precision at these stages, dPCR is still more reliable than any other molecular technique, such as qPCR.

12. How can I set the threshold by myself?

The Poisson calculation used in dPCR is based on treating each well as either negative or positive, which in turn is decided on by the fluorescent signal in the well. The QIAcuity Software Suite automatically sets the threshold at a certain level, but this can be manually adjusted. This manual adjustment is completely flexible, so you can set it wherever you like. Although it might seem arbitrary, it hardly differs from setting thresholds in imaging, flow cytometry, qPCR or when analyzing bands on gels for blotting or electrophoresis. Similarly, NGS relies on metrics to count a certain number of reads as true or noise. 

Setting a threshold in digital PCR for most assays with well-designed primers is relatively straightforward. Unless the separation of the clusters is poor, threshold setting doesn’t have to be done with high precision by the manual operator. Rather than having to hit the perfect spot, there’s a wider range in which all thresholds lead to similar numbers ± the 10% precision of the instrument.

13. What causes rain and how I can reduce it in my dPCR reaction?

In digital PCR data, clear separation between positive and negative results (signal-to-noise ratio) is crucial. Here “rain” describes the intermediate partitions that are not clearly positive or negative, making it challenging to classify them. These intermediate signals, appearing between the clearly defined “cloud” (positive) and “ground” (negative), are influenced by factors such as PCR efficiency, the presence of PCR inhibitors and off-target binding.

Adjustments in PCR parameters can help resolve these ambiguities.

For example:

• Increasing the annealing temperature to reduce off-target binding and clarify the separation between positive and negative signals: This approach helps determine whether the “rain” partitions are due to reduced efficiency or off-target effects
• Optimizing primer and probe concentrations
• Adjusting PCR cycle numbers
• Tweaking annealing and extension times

14. What does it mean if I see double positive bands on the 1D image?

It may mean you have non-specific amplifications. If your NTC is clean, you can try increasing the annealing temperature to make the reaction more stringent. If observed during multiplexing, this does not impact the concentration or fractional abundance of your targets.

15. What can lead to oversaturation?

Oversaturation of wells is caused by overloading of sample and is observed as the presence of only positive partitions. In this case, quantification is no longer accurate, as the statistical model used in dPCR assumes a random distribution and this requires a certain number of the partitions to be empty. We recommend at least 0.67% which corresponds to a maximum of five copies of target per partition.

To avoid oversaturation of a sample of unknown concentration, you can start by performing serial dilutions to see at which dilution you obtain the best signal-to-noise ratio. If multiplexing, it’s recommended to optimize each target individually (in a singleplex) and then multiplexing on a one-by-one basis (duplex, triplex, etc).

16. What are best practices for multiplexing?

Here are several tips for designing multiplex assays:

• Crosstalk is lower the further away the channels are from one another (so if possible, combine green and red as opposed to yellow and orange dyes)
• Choose assays with similar PCR conditions
• Multiplex targets with similar concentration in the template (c)DNA
• Reducing the amount of template can reduce interference of assays with one another
• Use primer and probe ratio of 2:1 – we recommend selecting 800 nM and 400 nM for singleplex; if you’re looking for more than a duplex, you could try lowering the concentrations
• Make sure the assays work fine in singleplex first
• Adding MgCl2, BSA, KCl, DMSO or extra polymerase can help with reaction efficiency
• Switching from a two-step to three-step process (separating annealing and extension) can increase reaction efficiency

17. Can two different reactions be set on the same plate with a different temperature profile? Is gradient dPCR possible?

The QIAcuity can run an essential gradient, but it cannot be programmed to run two or three different temperatures on the same plate. The gradient is always 12 different temperatures with a difference of roughly 1°C each. Hence, it’s possible to run a gradient, but not multiple reactions with different Tm.

18. What is the efficiency, productivity and difference in results when comparing two different approaches: PCR with probes vs PCR with primers alone?

When using well-designed primers with probes or with EvaGreen, you can expect very similar, if not identical results. That being said, using probes does have an edge when it comes to productivity since you can run only a single primer/EG-based assay per well, but multiple (currently up to five) probe assays in one well of the QIAcuity by using differently labeled probes. This way you can have 5x the information per well and sample

19. Can you give more details about the difference between the 8K and the 26K nanoplate?

The number of partitions per plate and the volume of sample that can be used are different in 8K and 26K plates. In the 26K plates, having slightly more partitions has a positive impact when trying to detect rare events (mutated DNA in cancer, for example). Being able to use a larger input amount of sample (up to 28 μl compared to only 8 μl in the 8.5K plates) also helps us detect rarer events.

19. Can you give more details about the difference between the 8K and the 26K nanoplate?

The number of partitions per plate and the volume of sample that can be used are different in 8K and 26K plates. In the 26K plates, having slightly more partitions has a positive impact when trying to detect rare events (mutated DNA in cancer, for example). Being able to use a larger input amount of sample (up to 28 μl compared to only 8 μl in the 8.5K plates) also helps us detect rarer events.

20. Can I use custom dyes? Can I use another mastermix? 

 A list of tested alternative dyes can be found in the QIAcuity Application Guide. You can use custom dyes as long as they fit the excitation/emission range of the QIAcuity detection channels (see Table 2). You should also check for crosstalk if using non-validated dyes. Our dPCR mastermixes are optimized for the low reaction volumes and the reaction conditions used in our nanoplates. We have tested other mixes and the performance is usually sub-optimal. Furthermore, the reference dye in our mixes is required for our software to do the analysis.