Amplification curves

Every reaction's fluorescence against cycle number with the fitted model and its quantification cycle, so a Cq is read from the curve it came off rather than trusted as a number.

What it doesCopy Anchor Link

An instrument hands back one number per well. That number is the end of a long inference — a baseline was subtracted, a threshold or a fit was chosen, and a cycle was read off where the curve crossed it. Two wells can report the same Cq from curves that look nothing alike, and the ones that went wrong are usually obvious the moment you see them.

This draws the reactions themselves: fluorescence against cycle, all wells on one axis, with the fitted curve over the observations and the quantification cycle marked where it was taken. A reaction that never leaves the baseline, one that starts high because of a primer-dimer, one whose plateau is a third of its neighbours' — each has a shape, and the shape is the diagnosis.

When you'd use thisCopy Anchor Link

• Before believing a fold change. A ΔΔCq is arithmetic on Cq values; if one of those values came off a reaction that never amplified properly, the arithmetic is still perfect and the answer is still wrong.
• When replicates disagree. Whether the spread is in the baseline, the exponential phase or the plateau tells you whether it is pipetting, inhibition, or a template difference.
• When a no-template control is not flat. Seeing where it lifts separates contamination from primer-dimer.
• When efficiency looks off. The slope through the exponential phase is what efficiency is measured from, and a curve that is not exponential where it is being fitted will not give an honest one.

What this computesCopy Anchor Link

• The fitted amplification model for each reaction, over the observed fluorescence.
• Cq, the quantification cycle — the cycle at which the reaction crosses the criterion, marked on each curve rather than only listed.
• Amplification efficiency derived from the fit, with the coefficient of determination for it.
• The baseline region each curve was corrected against.
• The per-reaction table underneath, so every drawn curve has its numbers beside it and the two cannot disagree.

What it does not doCopy Anchor Link

It shows what the reactions did; it does not decide which of them to keep. Excluding a well is a judgement about the experiment, and it is made explicitly rather than by a rule applied quietly on your behalf.