JOVANA
Library Glossary Getting Started Three Levels Fields How it works Mission
Join the mission
All guides

qPCR & RT-PCR: Measuring & Reading RNA

The previous guide showed PCR copying DNA a billionfold. Now two twists turn that copying into a ruler: reverse transcription lets PCR reach RNA, and real-time PCR watches the copies pile up so you can measure not just *whether* a target is there, but *how much*.

From yes/no to how-much

Plain PCR, as you met it in the last guide, is fundamentally a *presence* test. You run thirty-odd cycles, then look at the final tube on a gel: a band means your target sequence was there to be copied, no band means it was not. That is enormously useful — it answers "is this gene, this pathogen, this mutation present?" — but it throws away a question biologists ask constantly: *how much* was present to begin with? A cell with ten copies of a messenger and a cell with ten thousand copies can end the same PCR run looking nearly identical, because amplification runs until the ingredients are used up and then plateaus at roughly the same final amount whatever it started from.

This matters because so much of biology is a question of *amount*, not just identity. How strongly is a gene switched on in a tumour versus healthy tissue — its level of expression? How many copies of a virus are in a drop of blood — the viral load that tells a doctor whether a treatment is working? Two refinements turn PCR from a switch into a measuring instrument. The first lets it read RNA instead of DNA, so you can ask about gene activity at all. The second watches the copying happen live, so you can back-calculate the starting amount. This guide is about both.

RT-PCR: turning RNA back into DNA so PCR can read it

PCR's DNA polymerase only copies a *DNA* template — hand it RNA and nothing happens. So to measure RNA, which is where gene activity actually shows up, you need a translation step that is not the translation of the ribosome but a change of medium: copy the RNA into DNA first. The enzyme that does this is reverse transcriptase, and you have met it before — it is the same reverse transcriptase that retroviruses and retrotransposons use to write their RNA genomes back into DNA. It reads an RNA strand and lays down a complementary DNA strand, producing what is called cDNA (complementary DNA), a faithful DNA transcript of the original RNA message.

So RT-PCR is just two enzymes run in series: first reverse transcriptase turns your RNA into cDNA, then ordinary PCR amplifies the cDNA in the usual cycles of melting, annealing, and extension. One small but useful trick lives in how you prime the reverse-transcription step. If you want only finished, mature messenger RNAs, you can prime with a short stretch of T's — an oligo-dT primer — which base-pairs to the poly-A tail that mature mRNAs carry, so reverse transcription starts only on tailed messages. If you want every RNA in the sample, you use random primers that land everywhere. The choice quietly decides *which* RNAs you are about to measure.

A word on names, because they trip people up. "RT-PCR" means *reverse-transcription* PCR — the RNA-to-cDNA front end. "qPCR" (or "real-time PCR") means *quantitative* PCR — the live-monitoring back end of the next section. They are independent ideas that happen to be combined constantly: when you measure how much of an RNA virus is in blood, you run RT-qPCR, both at once. The shared letters "RT" are a genuine source of confusion, so keep the two meanings separate in your head even though the experiments usually fuse them.

qPCR: watching the copies pile up, in real time

The trick of quantitative PCR is to stop reading only the *final* tube and instead measure the product *after every single cycle*, while the reaction is still running. To do that you add a fluorescent reporter that glows in proportion to how much double-stranded DNA is present. The simplest is a dye, often SYBR Green, that lights up only when it slips into the groove of double-stranded DNA; with little product it is dark, and as copies accumulate the tube grows brighter. A machine reads that brightness at the end of each cycle, so instead of one endpoint you get a whole amplification curve — fluorescence climbing cycle by cycle.

A dye that binds *any* double-stranded DNA is cheap but undiscriminating — it would glow for the wrong product or for primers that have stuck to each other too. For a sequence-specific readout you can instead use a probe: a short DNA piece complementary to your exact target, carrying a fluorescent tag, that only lights up when it finds and binds its matching sequence. A clever design is the molecular beacon — a probe folded into a hairpin that keeps its glow switched off until it straightens out against the target. Either way the principle holds: more target copied this cycle means more light this cycle, and the machine is watching.

fluorescence
  ^
  |                                  _____________  <- plateau (reagents used up)
  |                              ___/
  |                          ___/   <- exponential rise (doubling each cycle)
  |                      ___/
  |- - - - - - - - -____/ - - - - - - - - - - - -  <- THRESHOLD line
  |________________/
  |   (background, target too rare to detect yet)
  +----------------|----------------------------> cycle number
                   Cq
  MORE starting target -> curve crosses threshold EARLIER -> SMALLER Cq
  LESS starting target -> curve crosses threshold LATER  -> LARGER Cq
Every qPCR well gives a fluorescence-versus-cycle curve. The cycle at which it crosses a fixed threshold is the Cq (or Ct). A sample with more target to begin with reaches the threshold in fewer cycles, so a smaller Cq means more starting material.

Now the key idea. Early on, the target is too rare for its glow to rise above background noise; eventually it grows abundant enough that the curve lifts off and climbs steeply, doubling each cycle. The decisive number is the threshold cycle — the cycle at which fluorescence first crosses a set line, written Cq (sometimes Ct). Here is why it measures starting amount: a tube that *started* with lots of target needs only a few doublings to cross the line, so it crosses *early*, at a small Cq. A tube that started with little target needs many more doublings, so it crosses *late*, at a large Cq. Because each cycle doubles the product, every one-unit drop in Cq corresponds to roughly twice as much starting material — a clean exponential ruler reading backward from when the glow appeared.

Reading a Cq honestly: relative, absolute, and digital

A Cq on its own is just a number; turning it into an *amount* takes a reference. There are two honest ways to do this. Relative quantification compares your gene of interest to a steadily expressed "housekeeping" gene measured in the same sample, then compares that ratio between conditions — answering "is this gene twice as active in tumour as in normal tissue?" without ever claiming an absolute count. Absolute quantification instead runs a standard curve: known amounts of target are amplified to map Cq onto real copy numbers, so an unknown sample's Cq can be read off as "so many copies per millilitre" — exactly how a gene-expression change or a viral load gets a hard figure.

There is a third way to count that sidesteps the curve entirely: digital PCR. Instead of one reaction in one tube, you split the sample across thousands of tiny droplets or wells — so few molecules per droplet that most get either zero copies or one. You amplify them all, then simply *count* how many droplets glowed positive versus stayed dark. With enough droplets and a dash of Poisson statistics (a way to account for the occasional droplet that caught two), that count is a direct, absolute tally of starting molecules — no standard curve, no assumption about amplification efficiency. Digital PCR trades the analog ruler of Cq for honest arithmetic, which makes it the method of choice when you need to detect a vanishingly rare variant against a huge background, such as a few tumour DNA fragments adrift in blood.

Why this turns biology into a measuring science

Step back and see what these refinements bought. Plain PCR gave biology a *detector*; RT-PCR and qPCR gave it a *meter*. Suddenly you could ask quantitative questions and get quantitative answers from a pinch of sample in an afternoon: how many fold does this gene rise when a cell is stressed; how does a patient's viral load fall day by day on a drug; is this faint band a real signal or contamination. The reason the world could test for a respiratory virus on an industrial scale was precisely RT-qPCR — reverse-transcribe the viral RNA, then count it by threshold cycle, millions of times a day.

Keep the limits in view as clearly as the powers. qPCR measures one or a handful of targets that you already know to look for — you have to design primers against a sequence you can name, so it cannot discover the unexpected. To survey *all* the RNA in a cell at once, without a prior list, you need the sequencing methods coming next in this rung, which read messages by the tens of thousands rather than counting one. And the cDNA step at the heart of RT-PCR is the same first move behind a cDNA library and behind RNA-seq: copy fragile RNA into stable DNA so the rest of the toolkit can read it. qPCR is the precise, targeted *measurement*; the upcoming guides are how you *read the letters themselves*, at genome scale.