digital PCR (dPCR)… The Future of qPCR?

In the past few years a “new” PCR technique has come into the scientific community, it is known as digital PCR (or dPCR, for us acronym loving scientists).  I thought it worth discussing some of the details of the utility of dPCR for those of you out there that are just getting into the real-time PCR world or those considering the transition to this new technique.

What is dPCR, it is basically a means to do hundreds, thousands, and possibly millions of tiny PCR amplifications from a single sample.  Why would you want to do this, might be your first question?  The answer, because you want to get quantitative numbers for the copies of template (target DNA or RNA) in a tube.  Certainly, if you divide a 1 microliter sample into 10,000 different reactions and that small volume contains 1000 copies of template, statistically you would expect 1 in 10 reactions to produce an amplification product.   That assumes that the designed assay being used to detect the target is actually capable of amplifying as little as one copy of template if it present in the reaction tube.  This assumption is one that is made for a variety of qPCR techniques, including dPCR.

So let us assume that the PCR that you are performing CAN actually detect a single copy of target DNA or RNA if it is present in the reaction tube.  The question then becomes how many actual PCRs do you really need to quantify the template material in the tube accurately?  First of all , it bears mentioning, that what we are talking about here is the number of amplifiable copies of the target DNA or cDNA (which is a DNA copy of RNA used as target template in a separate [though possibly in the same reaction tube just separated out biochemically either by time or temperature or both]) not total DNA in a tube.  An amplifiable copy of DNA (or cDNA) must be a co-linear stretch of sequence that contains both the forward and reverse primer binding sites.    Any breaks between the primers will give great primer binding (and use of PCR reagents, specifically dNTPs) and will result in altered primer sequences (due to primer extension without product completion). Also, they need to be complete primer binding sites as well, not merely portions of the primer binding sites as any altered primer binding sites will give poorer primer binding, thus reducing the overall sensitivity of the PCR to something likely to be less than the single copy detection capabilities which we are assuming.

If we add to the above assumption that the target that we will be detecting will be bound to some sort of fluorescent probe, and not simply detected using a non-specific double strand DNA binding dye, then the probe binding site(s) must also be invariant to maintain the assumed single copy detection capability of the assay.

How, in light of these assumptions, is merely dividing the sample into multiple reactions going to improve on these assumptions?  Simply put, it is not.

So then you must ask yourself, why do I want to have real-time PCR machine that is dedicated to performing digital PCR when I can have one that can do the whole gamut of real-time PCR applications AND with not many more additional reactions (and no additional assumptions) also do a pared down version of dPCR?

Realistically, are you really going to use dPCR on a daily basis to quantify the ‘exact’ number of copies of template in a sample?  Probably not, you might need this occasionally for new target standardization so that when you assign numbers to your qPCR standard curve you are putting in some sort of absolute number per dilution.  Then you simply use the built in software, that comes with virtually all real-time PCR machines to store the standard curve.  Until you run out of stock template, or you develop a new target assay you are not going to need that level of quantitative precision again.

So how are you going to do these calculations without a dedicated dPCR instrument?  There are a number of quantitative packages out there on the web and available commerically.  Here I have listed two, one is BioSTAT 2009 and another is QUALITY.  Of course, I would be remiss if I didn’t mention that our staff at Fluoresentric has been doing this type of template quantification for years and we offer both a DNA and an RNA stock template quantification service for existing assays and for those that you don’t even have yet.  You can find out more on our store or by contacting us at Fluoresentric ( or simply call 800.808.0490 x 100.

In the next installment of this topic I will go into the numerical details of dPCR and qPCR.