A qPCR standard curve is the ruler you measure every unknown against, and like any ruler, its errors do not stay put. They print themselves onto every quantity you read from it. The curve is built by serially diluting a known template across several orders of magnitude and reading the cycle threshold at each step, and the whole method rests on an assumption that is easy to state and easy to violate: that each dilution step is exactly the fold change you intended. Automate the dilutions well and the curve is a straight, efficient line you can trust. Automate them carelessly and you get a curve that looks plausible, quantifies confidently, and is wrong.
This is about the pipetting that stands behind the statistics. The math of a standard curve is unforgiving in a specific way, and understanding that way tells you exactly which transfers to protect.
Why the errors do not average out
The reason a standard curve punishes pipetting error is the log scale. Concentration spans orders of magnitude, so the x-axis is logarithmic, and the reaction reports cycle threshold, which is itself a log-domain quantity because each cycle doubles the product. A volume error that you would shrug off in a linear transfer becomes a systematic tilt in a log-linear fit.
The slope of the line is not decoration, it is the amplification efficiency. A slope of about minus 3.32 means the target doubles every cycle, one hundred percent efficiency, the ideal. If your dilutions are not the fold change you think, the slope moves away from that value and the efficiency you report is an artifact of pipetting rather than a property of the assay. Worse, the errors compound down the series: an error in the first dilution is carried into the second, the third, and every step after it, because each step is made from the one before. This is the defining hazard of serial dilution, that mistakes accumulate rather than average.
Mix at every step or the series drifts
The single most common cause of a bad automated curve is inadequate mixing between steps. A serial dilution transfers a portion of a well into diluent and expects that portion to represent the well faithfully, which is only true if the well is homogeneous when you sample it. Dispense concentrated template into diluent and it does not instantly disperse, especially if the template solution is denser or more viscous than the diluent, so a transfer taken too soon carries the wrong concentration forward and the error rides all the way down the series.
- Mix thoroughly after each dilution: several deliberate aspirate-and-dispense cycles, enough to homogenize the well before the next portion is drawn.
- Mix gently enough to avoid bubbles: the same energy that homogenizes a well entrains air, and a bubble in a qPCR well scatters the optical read, so the mixing has to be firm without being violent.
- Keep the mixing identical at every step: consistency down the series matters more than any single mix, because the fit assumes each step was made the same way.
A curve built with proper inter-step mixing sits straight on the plot. A curve built without it bends, and a bent curve does not merely look untidy, it assigns the wrong quantity to every unknown you read from that region.
Carryover is the other way a curve lies
Because a serial dilution spans a huge concentration range, the difference between the most concentrated standard and the most dilute is enormous, and any material carried from a high well into a low one distorts the dilute end where it matters most. A dilution transfer that reuses a tip drags a film of concentrated template into the next, more dilute step, inflating its apparent concentration and flattening the curve exactly where the curve is most sensitive.
The defense is fresh tips at every dilution step, which is not negotiable for a standard curve even though it costs tips. The same logic argues against techniques that shuttle small excesses between wells: what saves time on a routine plate falsifies a curve whose whole purpose is to be trusted at the dilute end. Spend the tips. The curve is the reference that validates everything else on the plate, so it is the last place to economize.
No-template controls and the shape of the plate
A standard curve is rarely alone on a plate, and how you lay it out interacts with the pipetting. The no-template control, the well with everything but template, tells you whether your setup is clean, and it is only meaningful if the setup could not have contaminated it. Fill controls with the same fresh-tip discipline and, where the deck allows, before the high-concentration standards rather than after, so the busiest, most contaminating transfers happen downstream of the well that is supposed to prove cleanliness. Replicates matter too: because the curve anchors quantification, running each standard in replicate and watching the spread tells you whether your dilutions are as reproducible as the fit assumes. Tight replicates are evidence the pipetting held; scattered ones are a warning the ruler is soft.
A standard curve does not fail loudly. It stays straight enough to believe and quantifies every unknown against dilutions that were never quite what you thought. The mixing and the fresh tips are what make the ruler real.
References
- S. A. Bustin, V. Benes, J. A. Garson, et al. The MIQE Guidelines: Minimum Information for Publication of Quantitative Real-Time PCR Experiments. Clinical Chemistry 55(4):611-622, 2009. gene-quantification.de/miqe-bustin-et-al-clin-chem-2009.pdf
- S. A. Bustin, et al. MIQE 2.0: Revision of the Minimum Information for Publication of Quantitative Real-Time PCR Experiments Guidelines. Clinical Chemistry 71(6):634, 2025. academic.oup.com/clinchem/article/71/6/634/8119148
- Real-Time PCR: An Essential Guide. Open-access reference on assay design, efficiency, and standard-curve interpretation. ncbi.nlm.nih.gov/pmc/articles/PMC3294352/