Quantitative PCR (qPCR)
An intercalating dye that fluoresces when bound to any double-stranded DNA, so it reports every product the reaction makes. Cheap, works with any primer pair and needs no probe design, but specificity has to be demonstrated on every plate with a melt curve.
Our ruleSimulations prepare you for the bench but don't replace hands-on practice.
SYBR Green
An intercalating dye that fluoresces when bound to any double-stranded DNA, so it reports every product the reaction makes. Cheap, works with any primer pair and needs no probe design, but specificity has to be demonstrated on every plate with a melt curve.
How it differs from the standard method
- Always finish with a melt curve: it is the only evidence that the signal came from your amplicon.
- Keep primers at the low end (100–300 nM each). Dimers are a dye problem far more than a probe problem.
- Dimers in the no-template control melt 8–15 °C below the amplicon; contamination melts at the amplicon's own temperature.
- One colour per well, so no multiplexing: a reference gene needs its own wells.
1What a Cq actually is
A real-time PCR instrument reads the fluorescence of every well once per cycle, so instead of one band at the end you get a curve per well. Early on the signal is flat: product is doubling, but there is far too little of it to see.
Baseline
The flat part the instrument averages and subtracts. Set it too late and it eats into a well that takes off early; too early and the subtraction is noisy.
Threshold
A horizontal line through the log-linear phase, above the noise and well below the plateau. Every well on the plate must be read against the same line.
Log-linear phase
The ten or so cycles where product really does double each cycle. This is the only region where a cycle number means a fixed factor of template.
Plateau
Reagents are running out and the curves converge. A threshold up here compresses real differences and gives late, unstable Cq values.
2The standard curve measures the assay, not the sample
Run a dilution series of a known template (a linearised plasmid, a gBlock, a pool of your own cDNA) and plot Cq against the log of the input. The points should fall on a straight line, and its slope is the assay's amplification efficiency: E = 10^(−1/slope) − 1.
3Why ΔΔCq is the fragile part
Relative quantification normalises your target to a reference gene, then compares conditions: ΔCq = Cq(target) − Cq(reference) in each sample, ΔΔCq = ΔCq(treated) − ΔCq(control), and fold change = 2^−ΔΔCq. That arithmetic hides two assumptions.
- Cq / Ct / Cp
- The fractional cycle where a well's curve crosses the threshold. Different manufacturers use different letters for the same number.
- ΔCq
- Target Cq minus reference-gene Cq within one sample. Normalising inside the sample is what cancels differences in how much cDNA went in.
- ΔΔCq
- The difference between two samples' ΔCq values. 2^−ΔΔCq is the fold change, assuming both assays are ~100% efficient.
- Amplification efficiency
- How close the reaction is to doubling every cycle, measured from a standard curve's slope: E = 10^(−1/slope) − 1.
- Reference gene
- A transcript assumed not to change with the treatment, used to normalise. An assumption to be tested, not a property of the gene.
- NTC
- No-template control: the mix with water in place of sample. In SYBR it also tells you whether the primers amplify each other.
- −RT control
- An RNA sample taken through the reverse-transcription step without the enzyme. It separates cDNA signal from genomic DNA contamination.
- LOD and LOQ
- The lowest input you can detect at all, and the lowest you can put a number on. They are different, and the second is higher.
4A dye or a probe
SYBR Green and its relatives fluoresce when bound to any double-stranded DNA. They are cheap, they work with any primer pair, and they report everything: your amplicon, primer dimers, and whatever mispriming happened at a low annealing temperature.
5Where qPCR stops being quantitative
At 100% efficiency, a single template molecule crosses the threshold at around cycle 36–37. So a Cq of 35 or later is a handful of molecules, and at that point the dominant source of variation is no longer your pipetting but Poisson sampling: whether three, five or eight molecules happened to land in that 10 µL.
Key terms
Practise these as flashcards →The flat early cycles the instrument subtracts from each well so that curves start at zero. A badly placed baseline shifts Cq without changing any chemistry.
The fluorescence level whose crossing defines Cq. It must sit in the log-linear phase, above the noise and below the plateau, and be the same for every well compared.
Fluorescence against temperature after cycling, used with a dsDNA dye to check that one product was made. Its derivative peaks at the product's Tm.
Primers annealing to each other and being extended. Short, melts low, invisible to a probe and fully visible to SYBR Green.
Cq against log input across a dilution series. Its slope gives amplification efficiency, its R² reports the dilutions, and its span is the assay's dynamic range.
Reading several targets in one well using probes with different fluorophores. Needs matched efficiencies and a check that the assays do not compete.
The published minimum information set for reporting a qPCR experiment: efficiency, controls, reference genes, replicate structure and analysis settings.
When to use it
- You want to know how much of a known target is there (transcript level, viral load, copy number) rather than just whether it is present.
- You are comparing the same target between conditions, and a fold change with a confidence interval is the answer you need.
- You need a sensitive yes/no on a defined sequence across many samples: qPCR detects single-digit copies and scales to hundreds of wells a day.
- You do not need an absolute number without a standard, and you do not need to resolve a rare variant at well under a per cent: digital PCR is the better tool for both.
Compared with alternatives
Endpoint PCR answers present or absent from a band. It cannot quantify, because by the time you look the reaction has plateaued and wells that started decades apart look the same.
Partitions the reaction and counts positives, giving an absolute concentration with no standard curve and far better precision for small differences and rare variants.
Prefer it when there is no thermocycler: LAMP and RPA run at one temperature and give a time-to-positive, at the price of much weaker quantification.
The same chemistry applied to RNA, with a reverse-transcription step in front. That step is where most of the variability in an expression measurement comes from.
Use a dye-based or spectrophotometric measurement when you need total nucleic acid. qPCR measures only amplifiable copies of one sequence, which is a different quantity.
Sub-techniques
Choosing one here switches the whole page, including the protocol and the lab, to it.
An intercalating dye that fluoresces when bound to any double-stranded DNA, so it reports every product the reaction makes. Cheap, works with any primer pair and needs no probe design, but specificity has to be demonstrated on every plate with a melt curve.
- Always finish with a melt curve: it is the only evidence that the signal came from your amplicon.
- Keep primers at the low end (100–300 nM each). Dimers are a dye problem far more than a probe problem.
- Dimers in the no-template control melt 8–15 °C below the amplicon; contamination melts at the amplicon's own temperature.
- One colour per well, so no multiplexing: a reference gene needs its own wells.