qPCR Controls Demystified: NTC, gDNA Contamination Check, and Interplate Calibrators

Your NTC just amplified at Cq 35 and your samples are at Cq 28. Do you trust the data, throw out the plate, or re-pipette the master mix? The answer depends on which control failed, what the melt curve says, and whether your reference genes were normalized to a contaminated baseline. This guide covers the three controls every qPCR run needs, how to design them so failures are interpretable, and what to do when one fires.

The three controls every qPCR run needs

NTC catches reagent and environmental contamination. NRT catches genomic DNA contamination in your RNA prep. The interplate calibrator (IPC) corrects run-to-run drift when an experiment spans multiple plates. MIQE-compliant studies report all three; in practice, NRT and IPC are skipped more often than they should be.

NTC (No-Template Control) — how to design and interpret

An NTC is a reaction with master mix and primers but nuclease-free water in place of template. Run at least two NTC wells per primer pair per plate. The expected result is no amplification — the curve stays flat, Cq is undetermined.

The practical pass criterion: NTC Cq is undetermined, OR NTC Cq is at least 5 cycles above the highest sample Cq for that target. The 5-cycle margin maps roughly to a 32-fold difference in starting template, which keeps a low-level NTC signal from dragging your ΔΔCt math.

When NTC fires, the question is what amplified, not just that something amplified. Run melt curve analysis on the NTC wells:

• NTC peak at a lower melt temperature than the sample peak → primer-dimer artifact. The primers self-anneal in the absence of template. Fixable: redesign primers, lower primer concentration to 100–200 nM, or raise annealing temperature. See interpreting melt curve peaks and primer dimers.
• NTC peak at the same melt temperature as the sample peak → cross-contamination. Reagent or environmental DNA matches your target. Discard the master mix aliquot, decontaminate pipettes, re-run.
• Both peaks present → both problems. Triage the contamination first; primer dimers can be tuned around once the reagents are clean.

Primer-dimer artifacts in the NTC are a primer design failure, not a contamination event. The fix is upstream — check your primers for self-complementarity and 3’ complementarity before running another plate. The primer design rules for qPCR post covers the dimer-avoidance checks.

NRT (No-Reverse-Transcriptase) — how to design and interpret

An NRT control is a mock reverse transcription with everything in the RT mix except the reverse transcriptase enzyme. Any amplification in the NRT means genomic DNA survived your DNase treatment (or you skipped DNase) and is being amplified by primers that don’t discriminate between cDNA and gDNA.

NRT matters most for two situations:

• Primers that don’t span an intron. If your forward and reverse primers sit in the same exon (or your gene is intronless), gDNA produces an amplicon indistinguishable from the cDNA product. Trace gDNA → false signal.
• Genes with processed pseudogenes. The pseudogene has the same exon sequence as the mRNA but exists as gDNA. Even intron-spanning primers can’t save you here.

Pass criterion: NRT shows no amplification, or NRT Cq is at least 5 cycles above the lowest sample Cq for that target. If NRT fires within 5 Cq of your samples, the gDNA contamination is high enough that DDCt fold-change estimates will be biased — treat the data as compromised until you re-prep RNA with a stronger DNase step.

The typical skip pattern: practitioners run NRT for “new” targets but assume their well-characterized reference genes (GAPDH, ACTB, 18S) don’t need it. Reference genes have processed pseudogenes too. ACTB has at least 6 known pseudogenes in the human genome. Skipping NRT on reference genes is the most common cause of normalization drift that nobody can explain.

Interplate calibrator (IPC) — how to design and interpret

An IPC is a single, well-characterized template (often a pooled cDNA from your control samples, or a synthetic standard) that you run on every plate of a multi-plate experiment. Its purpose: detect and correct plate-to-plate Cq drift caused by master mix lot changes, instrument variation, or threshold-setting differences across runs.

You need an IPC any time your experiment spans more than one plate. If you can fit all samples and replicates on one plate, skip it; if you can’t, skipping it is the silent killer of multi-plate qPCR studies.

Design notes: pick an IPC template at moderate abundance (Cq around 22–28 in your assay) so it’s sensitive to drift without being efficiency-limited. Run the IPC in triplicate on every plate. Use the same IPC for every target gene on a plate — the correction is per-plate, not per-target.

gDNA contamination check — how to verify

NRT detects gDNA contamination in the RNA → cDNA conversion. The complementary check is at primer design time: intron-spanning primers can’t produce a short amplicon from gDNA because the intron makes the gDNA template too long for the extension time. If your primers span an intron of >1 kb, gDNA contamination is geometrically suppressed even without perfect DNase digestion.

Two-prong defense:

• Design intron-spanning primers when possible. Verify the intron position in NCBI’s gene structure viewer. Confirm with a melt curve — gDNA amplicons (when they appear at all) typically show a different melt profile than the cDNA target.
• DNase-treat your RNA prep, then run NRT. If your gene has a pseudogene or you can’t span an intron, DNase treatment plus NRT verification is the only path. Use a column-based RNA cleanup post-DNase — residual DNase carries into the RT step and degrades cDNA.

Interpreting failed controls — decision tree

The decision tree is intentionally rule-based. Eyeballing “the NTC looks fine” on a curve display is how plates that should have been thrown away end up in publications.

Common mistakes

• Skipping NRT for reference genes because they’re “well-known.” GAPDH, ACTB, and 18S all have pseudogenes or paralogs. A reference gene that picks up gDNA signal in some samples and not others creates a normalization artifact that looks like real biology. Run NRT for every target, including normalizers. See validating reference genes with GeNorm and NormFinder for the stability layer that catches this downstream.
• Mistaking primer dimers for contamination. If your NTC fires but the melt peak is 4–8°C below the sample peak, you don’t have a contamination problem — you have a primer design problem. Different fix, different urgency. Check the melt curve before nuking the lab.
• Cross-contamination from pipetting. The most common contamination source is the practitioner’s own pipettor. Use filter tips. Set up NTC wells before sample wells, with a fresh tip for each well. If your NTC fails on a single well but not adjacent NTCs, suspect a tip change you forgot to make.
• Treating the IPC as optional for two-plate experiments. Two plates is enough for plate-to-plate drift to matter. The cost of an IPC is three wells per plate. The cost of skipping it is a re-run when reviewers ask why your fold changes don’t replicate.
• Reporting “NTC clean” without a Cq threshold. MIQE compliance means stating the criterion. “NTC was undetermined or ≥ 5 Cq above lowest sample Cq for all targets” is reportable; “NTC was clean” isn’t.

Controls are the boring part of qPCR until they fail, at which point they’re the only thing standing between you and a retracted figure. Design them in at the plate-layout stage, not as an afterthought when reviewers ask. If you want help interpreting control failures across a real dataset, AnnealIQ’s QC engine reads the controls alongside your samples and surfaces what to do about each one.