In our last blog post, we promised you more rarely asked questions (RAQs) on dPCR, and now we’re delivering. Dr. Jan Rohde carefully collected more rare gems during the second part of his “Ask me anything again: Digital PCR edition” webinar, and you can find his famously easy-to-understand explanations here.
1. How does the partitioning process in digital PCR affect the accuracy and sensitivity of the assay?
In most digital PCR systems, calculations are based on Poisson statistics. This means a random distribution of template DNA/RNA molecules across the partitions is necessary to allow accurate and sensitive detection of the targets. Because of this, you need to thoroughly mix the reaction volume before partitioning. For the QIAcuity, this mixing is done manually in a pre-plate or robotically in the nanoplate itself.
Additionally, long DNA/RNA molecules are “sticky” as they tend to wind around one another and not homogenize well. It’s recommended to digest longer nucleic acids down to a size of 20,000 base pairs or less. Finally, digital PCR mixes have to fill each partition equally or risk sub-optimum performance.
2. What are the limitations of digital PCR in terms of template copy number and dynamic range?
The dynamic range of dPCR is generally about 5 log values, including measurements at non-optimal statistical ranges. Template copy numbers depend on the number of partitions and the volume of dPCR mix in the partitions. Generally, if you aim for precise measurement, we suggest targeting about 0.5—3 copies of your target per partition. A range between 0.05 and 5 still works, but if you have a higher or lower number of molecules than expected, you could lose precision.
For our nanoplate partition sizes, 0.5 to 3 copies translate to roughly 4250 and 25500 molecules for the 8.5K Nanoplate and 13000 and 78000 molecules in the measured volume for the 26K nanoplates. As both plates have a little access volume, you can actually use more DNA than that and still obtain accurate measurements. You can find the exact numbers in our application manual. Please note that you can detect significantly lower amounts of DNA (LOD depends on the assay, but I have seen as low as 6-10 molecules), but the precision is going to decrease.
3. How do different digital PCR platforms (droplet-based vs. chip/plate-based) compare in terms of reproducibility and data consistency?
The QIAcuity uses nanoplates instead of droplets. Nanoplates have the advantage of being less sensitive to impurities in the reaction volume (such as detergents). We can also check the actual size for the partitions in each well to account for small differences in volumes per partition between wells. Droplet size, on the other hand, has been described to vary by 2—20% depending on which literature you’re looking at and what mixes are used. When it comes to sensitivity and specificity, ddPCR and nanoplate dPCR are very comparable for the majority of assays. For a more comprehensive comparison, visit our bench guide on dPCR technologies.
4. What are the potential sources of error in digital PCR and how can they be mitigated?
Honestly, this question is too broad to answer. I mean you can also ask what can go wrong when making a sandwich, which might be not using any bread and instead boiling water to make pasta. If we re-phrase and say common sources of error, I can give meaningful answers.
There are many possible sources of error in digital PCR but I have found some common ones in my years of support experience. Mixing of the reaction volume is of key importance and is something that can often be overlooked. I usually recommend vortexing for 5—30 seconds, but you can also pipette up and down with 80% of the volume at least 10 times. Anything less than that is not thorough enough. PCR inhibition, while less of an issue than in qPCR, can still impact digital PCR results, for example, when carrying over ethanol from the DNA purification. The filling of the wells can also be a bit tricky the first few times you use the system. You can’t spin down the nanoplates, so you need to deposit the reaction volume at the bottom of the nanowells. I usually move the pipette to the wall of the well and slide down into the corner and then deposit the reaction mix. I’m also careful to only push in the pipette pin until the first resistance to avoid introducing any bubbles.
5. How does the presence of inhibitors in a sample affect digital PCR performance and what strategies can be employed to overcome this?
Generally speaking, inhibitors reduce PCR efficiency, which in dPCR can range from a slight reduction of fluorescent signal to a complete loss of any signal, depending on how bad the inhibition is. Using high quality DNA/RNA isolation kits can reduce the amount of inhibitors in your samples. Furthermore, our OneStep mixes are particularly resistant to inhibitors. The Q-solution Kit, initially developed for difficult-to-access DNA, is also effective at countering some inhibitors like ethanol.
6. What are the best practices for designing assays to minimize false positives and negatives in digital PCR?
The best practices for assay design can be found in our application manual. Overall, the best practices are identical to assay design guidelines for qPCR, where you try to maximize the PCR efficiency to obtain a good signal. False positives are caused by off-target binding, which you can check by blasting your primers in silico. They can also be produced by primer dimers interacting with probes. False negatives are usually caused by low PCR efficiency. In these cases, you fail to amplify every single molecule in the sample. This is usually less of an issue, as most primers designed by contemporary software will amplify the target successfully. Low PCR efficiency becomes a greater problem if the sample is not very clean. In such cases, PCR inhibition can lead to loss of molecule count.
7. How does digital PCR handle complex samples with high background noise or multiple targets?
Digital PCR is very useful for picking up a small number of target molecules in a large amount of different DNA molecules. By splitting the template DNA into thousands of partitions, you can dilute out the more prevalent rest DNA that has a smaller impact on the PCR reaction, so you get better resolution above the background noise. The same holds true for multiple targets. Analyzing multiple targets is easier to achieve in dPCR, as in each partition, you’re likely to only have one or maybe a few target molecules instead of the thousands or millions in a qPCR reaction.
8. What are the advancements in digital PCR for single-cell analysis and how do they impact the interpretation of results?
We have several application notes showing that you can sort a single cell, lyse it and then analyse the contents using digital PCR. Again, the ability to precisely and reliably detect molecules comes in handy here whether you want to measure the hundreds of copies of mitochondrial DNA from one cell or a low number of microRNA molecules.
9. How do digital PCR results compare with other quantitative PCR methods in terms of clinical utility and regulatory acceptance?
This question has to be taken on a case-by-case basis. There are numerous qPCR assays that have been validated for use in diagnostics and only a small number of validated digital PCR assays due to the technique gaining more widespread use only in recent years. However, due to the increased precision and better ability to detect a small number of target molecules in high background noise, dPCR should, theoretically, have higher clinical utility than qPCR. One such case could be analyzing liquid biopsies to detect cancer or biomarkers for various diseases. Regulatory bodies aren’t always aware of the advantages of dPCR, but in certain cases have already transferred to dPCR from the current standard qPCR.
Dr. Jan Rohde loves answering these questions. Let’s keep him happy. Send him of your questions on social media and follow our social media channels and the blog for future Q&A editions.