TaqMan Probe Design Rules: Why the Probe Must Sit Above the Primers
7 min read · Updated November 11, 2026
A qPCR primer pair and a qPCR probe are designed against different constraints, and the one that ties them together is a temperature gap. The probe has to be fully annealed before the primers extend, or the polymerase will displace it instead of cutting it, and no signal appears.
That single requirement generates most of the probe rules people half-remember, including the length, the GC window and the ban on a 5' guanine.
The five rules, and where each comes from
- Length 20–30 nt. Short enough to synthesise cleanly and to discriminate a single mismatch, long enough to reach a Tm above the primers at a workable GC.
- Tm 6–12 °C above the primer Tm, with 8 °C as the target. This is the rule the others serve: the probe must be bound first.
- No G at the 5' end. A 5' guanine quenches the reporter even after the probe is cleaved, so the signal stays low no matter how well the assay amplifies.
- No run of four or more identical bases — three is the longest the designer accepts. Runs make synthesis fail more often and make the probe harder to discriminate.
- Inside the amplicon and clear of both primers. A probe that overlaps a primer competes with it for the same bases.
Why the strand choice is not arbitrary
A probe can be written against either strand. The convention worth following is to put it on whichever strand has more C than G in that window, because it is the C content that sets how well the probe behaves after cleavage — and the 5' base of the chosen strand then decides whether the reporter is quenched.
That is why the two rules interact, and the second assay from the same run shows it happening. Its window on the plus strand is GCCAACACTTGTCACTACTTTCTCTTATGG — nine C against four G, so the plus strand is preferred on composition, but it starts with G. The designer therefore takes the reverse complement, CCATAAGAGAAAGTAGTGACAAGTGTTGGC, which starts with C. Same window, same base pairs, opposite strand, and the quenching problem is gone.
Worked: one assay, five numbers
Run the qPCR assay designer on a 400 bp GFP template and it returns five assays, three of them with a probe placed. The first: forward GCACTACTGGAAAACTACCTG, 21 nt, Tm 59.1 °C, 47.6% GC; reverse GGCATGGCACTCTTGAAAA, 19 nt, Tm 59.4 °C, 47.4% GC. The primers are within 0.3 °C of each other, and the product is 124 bp.
The probe is CATGGCCAACACTTGTCACTACTTTCTCT, 29 nt, Tm 67.5 °C at 44.8% GC, on the plus strand. The gap that matters is measured against the WARMER primer: 67.5 − 59.4 = 8.1, reported as tmAbovePrimers 8, with the probe starting 3 bp past the forward primer's 3' end.
The 3 bp gap is not a defect. The probe sits just 3' of the forward primer, which is the arrangement the rules ask for: the polymerase runs into the hybridised probe from behind, and the closer it starts, the sooner the first cut happens. A probe placed 30 bp downstream works, and the signal arrives a little later in the cycle.
The 124 bp amplicon is the other number worth noticing, because a general primer designer would happily return a 600 bp product. A short product doubles cleanly every cycle; a long one is the commonest reason an assay's measured efficiency lands near 80% instead of near 100%.
Two of the five assays come back with no probe at all. That is the honest answer rather than a failure: the window between those primers holds no oligo that clears the length, GC, run and Tm-gap rules together. A pair worth keeping can still be a pair with no legal probe, and the designer reports it instead of relaxing a rule to fill the slot.
Checking the probe you already have
The same arithmetic runs in reverse on a probe someone else designed. Paste the oligo into a Tm calculator set to your own salt and primer concentration and compare it with the warmer of the two primers under the same conditions. The probe above scores 67.5 °C on its own, matching the designer's figure, so the two agree about the same molecule — which is the point of checking.
Do not compare a number computed at one salt concentration with a number computed at another. A published probe Tm quoted at 50 mM Na⁺ and a primer Tm computed in your own 3 mM Mg²⁺ buffer differ by several degrees before any biology enters, and the comparison will read as a design fault when it is a units fault.
What the design does not decide
- Reporter and quencher chemistry, and whether the pair is spectrally compatible with the other channels on your instrument.
- Whether the probe spans an exon junction. For RNA work that is a design choice made against the transcript, not against the amplicon.
- Whether the target region is free of SNPs in your samples. A probe sitting on a common variant fails only in the samples that carry it, which looks like variable efficiency.
- Efficiency itself. That is measured from a dilution series after the fact, not predicted from the sequence.
FAQ
How much higher should a TaqMan probe's Tm be than the primers?
6 to 12 °C above the warmer primer, with about 8 °C as the target. The probe must be fully bound before the primers extend so that the polymerase cuts it rather than displacing it.
Why can a TaqMan probe not start with G?
A guanine at the 5' end quenches the reporter even after cleavage, so the well stays dark however well the reaction amplifies. Choosing the other strand usually solves it, because the 5' base flips with the strand.
How long should a qPCR amplicon be?
70 to 150 bp. A short product doubles cleanly every cycle; a 600 bp product is the commonest reason an assay's measured efficiency lands near 80% instead of near 100%. The designer above returns 124 bp with the probe 3 bp from the forward primer.
Can I use the same primers for SYBR and for a probe assay?
Often yes, but the probe has to fit between them with 6–12 °C of Tm margin, so a primer pair designed without that headroom may leave no legal probe window. Design the probe and the primers together rather than retrofitting.
Related references
Related tools
Ranked PCR primer pairs from a template, with Tm, GC and dimer checks.
Primer Tm, GC content and molecular weight.
Nearest-neighbor Tm, ΔG, hairpins and dimers for oligos and primer pairs.
Fold change from Ct values, with propagated error and efficiency-matched method.