Introduction to Digital PCR (dPCR)
Q&A with Marcia Slater, Senior Technical Specialist for real-time PCR and digital PCR at Thermo Fisher Scientific
The details of what make digital PCR (dPCR) different from real-time, or quantitative PCR (qPCR) are relatively simple but not always explained very well. Likewise, it’s not always clear which use cases are a good fit for dPCR, and which others simply don’t require the power of dPCR. The power of digital PCR is real, if you understand it.
In this conversation Marcia Slater, a self-described “PCR guru”, explains digital PCR and its power. She covers the basic differences between dPCR and qPCR and then delves into the details of where dPCR derives its power and where it shines. With over 20 years’ experience in helping customers troubleshoot PCR, Marcia makes it easy to understand key terms and concepts related to dPCR, including:
- Poisson statistics
- Statistical power and confidence intervals
- Controls and false negatives vs. true negatives
- Dead volume
- Dynamic range
To learn more, listen to the Absolute Gene-ius dPCR Podcast Episode at www.thermofisher.com/absolutegeneius
Transcript from Absolute Gene-ius episode “Insights from a PCR whisperer”. This transcript has been edited for clarity and brevity.
The Power of dPCR
Question: So, talking about digital PCR, we want to dive in a little bit deeper into that and do something unique on this episode, especially since we are we are joined by such an amazing expert with digital PCR. Talk a little bit about, what is it? What is digital PCR? How is it different than real-time PCR? There’s a lot of buzz around this technology in labs and in the market. And so, we think it’d be interesting to kind of dive into that technology.
Marcia Slater:What digital PCR does is, it takes a reaction and breaks it up into many, many sub-reactions. In those early days, you were looking at hundreds, hopefully thousands.
Now we break a reaction up into tens of thousands of little sub-reactions. But then underneath of that breakout, the question we ask in each of the sub-reactions is simply if it’s positive or negative. So, each of the sub-reactions has the most simplistic of questions that you’re using, but by compiling all of those yes and no answers over, say 20,000 sub-reactions, 20,000 chambers, we get a very accurate and precise absolute quantity measurement of that target.
That’s one of the main reasons why people use digital PCR: to get that absolute quantity.
You also tend to have more statistical power when you use digital PCR, so that’s a popular reason why it’s being used. With how this works, beyond just the positives or negatives, if you get a chamber that’s positive, one of the questions becomes, “was it only one molecule in there? Or was it more than one?”
Poisson Statistics for Precision
Slater (cont.): With digital PCR, when you separate your material out into these sub-reactions, each of the targets will randomly go into any of the sub-reactions, and that probability that it will land in a well follows a Poisson distribution.
So besides this very simplistic plus or minus answer, there’s also the Poisson statistics that are the foundation of digital PCR. It’s a little more complex than just adding up the number of positives that you see, it’s applying this correction factor for the possibility and probability that a chamber had more than one.
Understanding the Process
Question:You talk about breaking it out into thousands of these sub-reactions, right? Or breaking it into these different chambers? What does that look like? How is that different in digital PCR? What actually happens to that sample?
Marcia Slater:Yeah, you’re right. In qPCR, you have that one single tube, and you get a Ct or a Cq value out of it. It’s single answer that comes out of that tube for real-time PCR.
In digital PCR, you take what amounts to the essentially the same tube, but instead of getting that Ct value, you spread it out before the PCR reaction to all the sub-reactions. If you had, say a hundred positives over with real-time PCR, you would have gotten a certain Ct value. When you spread that out into these chambers of digital PCR, you might get potentially as many as a hundred positives out of, say, 20,000. Now it might be lower, because just by random assortment, some of the chambers may have gotten two or three instead of only one. But that’s really the difference in how it works. And then we already talked a little bit about Poisson.
Counting Positives and Negatives
Slater (cont.): But one of the interesting things, once you do that spreading out of the sample, we count up the positives, and we count up the negatives. It’s actually the negatives that are the most important for getting your quantity. Because when we’re modeling that data into the Poisson distribution, we only know two things, positive or negative. If we plot out bar graphs of Poisson of the number that are there, the only group that we know with certainty is the number of negatives.
We know that that group is discrete, they’re just negatives, that’s that bar graph. But then the positives is essentially the sum of all the ones that have one, two, or more targets in that, so we don’t know any of those single bars with absolute certainty, just the number of negatives. The number of negatives anchors us in Poisson. And then once we fit that data to the Poisson model, the average number of copies per chamber can be computed.
That’s called the lambda in digital PCR and once you have that, then it gets really easy.
Once you know the average, we know the number of chambers that got filled. And with physical wells, we know the volume of those wells, so that lets us get to the copies per microliter. So instead of getting a Ct value, which is a timing measurement, we get an absolute quantity of copies per microliter. Or if you want to be really technically correct, it’s the molecules per microliter of that target.
Applications of Digital PCR
Question: In practice, what are common applications then that digital PCR might make sense to have as a as a tool?
Marcia Slater: There’s a lot of different applications that can utilize digital PCR.
Most commonly, I see people who are doing just straight up absolute quantification of their material. That could be for gene expression studies. It could be for a viral titer. We talked about absolute standard curves. One of the uses for digital PCR is to quantify standards to be used for real-time PCR absolute standard curves. With real-time PCR, absolute quantification with a standard curve, your quantity is only as good as how well you’ve quantified those standards. A lot of times standards that come from a vendor aren’t that precisely quantified.
Using digital PCR can give you a better quality absolute standard curve if you’re doing real-time absolute standard curves. There are other applications as well, such as those that deal with quantifying a single base change. That’s something that’s really hard to do with real-time PCR. It’s possible, but it’s often not that straightforward, but with digital PCR, we have the ability to quantify just single-base changes. So, with that there are a lot of people who are doing things like looking for driver mutations in a cancer sample. And then the same exact chemistry that allows you to look at say, an oncology mutation, can also be used to look at variants in microbes. If you were genotyping a microbe, you need a percentage of how much of that base is there to represent the different strains. Digital PCR will let you quantify that.
When to Choose qPCR vs. dPCR
Question: If somebody is looking at trying to decide qPCR versus dPCR, one of the best things they can do is just look at “What is it I’m trying to achieve?” And that can start to help them along the way of deciding between the two?
Marcia Slater: Yep, that is correct. With qPCR, you tend to have a wider dynamic range. If you have no idea what you’ve got, want to just throw it in there, get a quick answer, get an inexpensive answer, qPCR is a great way to go. It’s when you want to get more statistical power that I would move to digital. Digital will have a narrower dynamic range than qPCR, so you need to be aware of that.
If you have a really abundant target, you may need to dilute that in order to bring it down into the digital PCR sweet spot. But if you hit that sweet spot, you will get precision levels that qPCR just can’t even dream of.
Exploring dPCR Technologies
Question: Very cool. Well, Marsha, I want to talk about your love of technology and how these are built. Focusing on digital PCR, what technologies are out there? Are there different approaches for digital PCR, and maybe some advantages and disadvantages of those?
Marcia Slater:At Thermo Fisher Scientific, we believe, all of our generations of digital PCR have used physical wells. That’s actually how digital PCR started. In the early days, it was 96-well plates. And then the second generation of digital PCR were also primarily plate-based in the open array and other cartridges that had little reaction chambers.
Later came droplets. Droplets are aqueous droplets in an oil background. The concern that I have with the droplets is, “Are they absolutely uniform?” That’s one of the things to be aware of. Also, they’re very fragile.
As you handle droplets, you can actually lose droplets if they merge back together. With physical wells, we know precisely how big that well is. One of the really clever things that’s done with the chemistry is to use a dye for quality control.
Using ROX for Quality Control
Slater (cont.): With our Absolute Q system, we use ROX for quality control. We look at that well and if that well does not have the target signal that is a candidate negative well. Then we look at the ROX. If the ROX is there and it’s at the proper amount of ROX, then we know that that is a correctly filled well, that everything is there, and the only thing that’s missing is the target. That’s a true negative well.
But if the ROX isn’t there, then that’s an empty well. If it’s not in the right quantity, it’s an improperly filled well, those are removed, those are false negatives. I mentioned before that negatives are the most important thing when we’re counting up the positives and negatives. We have to have that number be perfect.
Dead Volume in Experiments
Question: Interesting. And one more buzzword that that I’ve heard going around is dead volume, right? Is there any insight you can give on dead volume? And, if somebody is looking to get into digital PCR, why is dead volume such a critical thing to look at?
Marcia Slater: Yeah, so dead volume, the importance is going to vary depending on your application.
If I had boatloads of my target material, if I was maybe manufacturing a nucleic acid and I had liters and liters, dead volumes not going to be a big deal to me. Because what dead volume is, it’s the amount of your material that’s just essentially lost in the plumbing of that experiment. It’s gone. It’s like you’ve thrown it away.
If you throw a portion of your material away, and you have boatloads of it, then you don’t care if you’ve lost a few microliters here or there. But if I’m one of those folks that’s doing that fine needle aspirate, or liquid biopsy, or something else, that’s a really precious experiment. I once worked with a person who was doing an experiment from a space shuttle. You’re not getting another chance at that space shuttle experiment. That’s a situation where if you lose any amount of that and it’s gone, that’s it, you can’t repeat it. For those experiments, dead volume is critically important. It’s important to have the lowest possible dead volume if you’re working with a precious sample like that.
The Future of dPCR
Question: Awesome, we’re talking about cool things and space stations being one of them. Are there any really interesting things that you’ve seen digital PCR be used for?
Marcia Slater: It’s becoming a big thing with cell and gene therapy. There’s a new kind of up-and-coming application that people will refer to as molecule integrity. That’s something that you actually can’t even do with qPCR. You may have noticed that earlier, I said that if you want to be totally technically correct, that digital PCR measures molecules.
Detecting Construct Integrity
Slater (cont.): For most applications, molecules and copies are the same thing. But imagine if you had a construct, and you not only wanted to know how much of that construct you had, but is the construct intact, versus getting degraded? With the machines today, you can do a multicolor analysis, they’re enabled for high levels of multiplexing.
Slater (cont.): You can put a different color probe assay spanning that molecule of interest, and then the question you can ask is, “Are all four colors together in my sub-reaction in my chamber?” If they’re together, then that implies that they were on the same molecule; think of those as being tied together with a rope and they’re all being pulled together into the same micro-chamber. If instead, you only have a subset of colors, so if you did a four color one, and only two are there, well, then that molecule is not there in its entirety. It must have been cut apart; otherwise, all four colors would be there.
So digital PCR really allows this kind of up-and-coming application to see if what you’re doing is intact, or if it’s been degraded.
Question: What are the practical implications of that? Why would somebody want to see if their molecule is intact?
Marcia Slater: Well, so one, they may have actually manufactured a construct that they’re moving into a cell line and it is important to know, “Is the whole thing there as I designed it?” Otherwise, I’d be wasting my time on a cell line that doesn’t have the proper construct in it. Another area where I see people using this is infectious disease. One of the questions these folks might ask is, “Is that whole viral genome present? Or has it been degraded in the reservoir in the cell?” So digital PCR gives us a tool to be able to answer those questions.
To get more information and hear Marcia Slater’s career advice, listen to the Absolute Gene-ius dPCR Podcast Episode at www.thermofisher.com/absolutegeneius
Learn more about applications of digital PCR at https://www.thermofisher.com/us/en/home/life-science/pcr/digital-pcr/resources.html