Lab TechniqueAdvanced~15 min

PCR & Primer Design Lab

Design primers, compute Tm properly, and run the thermocycler

Design forward and reverse primers against a template, compute melting temperature by both the Wallace rule and nearest-neighbor thermodynamics, then run the amplification.

The takeaway

The reverse primer is the reverse complement of the template — the single most common conceptual error in molecular biology, and one you can only unlearn by doing it.

Reverse complementNearest-neighbor TmGC clampPrimer dimersAnnealing temperatureExponential amplification

1 · Template DNA

600 bp, ~62% GC. The workhorse fluorescent reporter — well-behaved primers.

600 bp61.8% GCValid ACGT

2 · Primers

Primer length is optimised (18–30 nt) to land the nearest-neighbour Tm near 60°C with a clean 3' GC clamp and minimal self-structure.

Forward primer

20 nt
5'-CCATCCTGGTCGAGCTGGAC-3'
GC
65%
Tm Wallace
66°
4(G+C)+2(A+T)
Tm nearest-nb
58.7°
SantaLucia
3' clamp
2/3
G/C in last 3
ΔH°
-157.9
kcal · mol⁻¹
ΔS° (salt-corr.)
-442.8
cal · mol⁻¹ · K⁻¹

Reverse primer

21 nt
5'-CCTCCTTGAAGTCGATGCCCT-3'
GC
57%
Tm Wallace
66°
4(G+C)+2(A+T)
Tm nearest-nb
58.1°
SantaLucia
3' clamp
2/3
G/C in last 3
ΔH°
-162.4
kcal · mol⁻¹
ΔS° (salt-corr.)
-457.3
cal · mol⁻¹ · K⁻¹

Why two melting temperatures?

The Wallace rule Tm = 4(G+C) + 2(A+T) treats every base as an independent contributor. It is a back-of-the-envelope rule from the 1970s: fine for a 17-mer probe in 1 M salt, and wrong everywhere else. It cannot see sequence order at all — GCGCGCGC and GGGGCCCC get the identical Tm from Wallace, even though they melt several degrees apart.

It also ignores salt and primer concentration entirely, so it drifts badly for primers longer than ~20 nt.

Nearest-neighbour (SantaLucia 1998) models duplex stability as the sum of stacking interactions between adjacent base pairs. Each of the 10 unique dinucleotide steps carries a measured ΔH° and ΔS°; add helix initiation terms, correct ΔS° for salt, and solve

Tm = ΔH° / (ΔS° + R·ln(C_T/4)) − 273.15

This is thermodynamics rather than arithmetic, so it responds to sequence context, [Na⁺] and primer concentration. It is what every real design tool uses, and it is what sets the annealing temperature you should actually type into the machine.

Wallace ΔTm vs NN (F): 7.3 °CWallace ΔTm vs NN (R): 7.9 °CTry the GC-rich template — Wallace overshoots badly.

3 · Primer pair validation

Usable, with 3 cautions
Fwd length

20 nt — the sweet spot is 18–25 nt: long enough to be unique in a genome, short enough to anneal fast.

Fwd GC content

65.0% GC — aim for 40–60%. Too low and the primer is unstable; too high and it sticks non-specifically.

Fwd Tm (NN)

58.7 °C — a 55–65 °C Tm lets you anneal near 60 °C, where Taq is specific and still fast.

Fwd 3' GC clamp

2 G/C in the last 3 bases — a proper clamp: stable 3' end without being over-sticky.

Fwd hairpin

Best stem = 3 bp over a 3 nt loop (score 5). A stem of ≥4 bp folds the primer back on itself instead of onto the template.

Fwd self-dimer

Longest self-complementary run = 4 bp (score 20), involving a 3' end. Two copies of the primer annealing to each other burn primer and template each other.

Rev length

21 nt — the sweet spot is 18–25 nt: long enough to be unique in a genome, short enough to anneal fast.

Rev GC content

57.1% GC — aim for 40–60%. Too low and the primer is unstable; too high and it sticks non-specifically.

Rev Tm (NN)

58.1 °C — a 55–65 °C Tm lets you anneal near 60 °C, where Taq is specific and still fast.

Rev 3' GC clamp

2 G/C in the last 3 bases — a proper clamp: stable 3' end without being over-sticky.

Rev hairpin

Best stem = 2 bp over a 6 nt loop (score 3). A stem of ≥4 bp folds the primer back on itself instead of onto the template.

Rev self-dimer

Longest self-complementary run = 3 bp (score 12). Two copies of the primer annealing to each other burn primer and template each other.

ΔTm (F vs R)

ΔTm = 0.6 °C. Above 5 °C the two primers cannot share one annealing temperature — the "hot" primer mis-primes at the Ta that the "cold" one needs.

Primer-dimer (F×R)

Longest F×R complementary run = 4 bp (score 16), involving a 3' end. A 3'-overlapping pair extends each other into a short dimer that eats the reaction and shows as a smear below ~100 bp on a gel.

Annealing temp (Ta)
53.1 °C
≈ lower Tm − 5 °C
Amplicon length
360 bp
Fwd 5' end → Rev binding 3' end
Extension time
22 s
Taq ≈ 1 kb / min

Template map

Forward primer siteReverse primer binding siteAmplicon
1ATGGTGAGCAAGGGCGAGGAGCTGTTCACCGGGGTGGTGCCCATCCTGGTCGAGCTGGAC
61GGCGACGTAAACGGCCACAAGTTCAGCGTGTCCGGCGAGGGCGAGGGCGATGCCACCTAC
121GGCAAGCTGACCCTGAAGTTCATCTGCACCACCGGCAAGCTGCCCGTGCCCTGGCCCACC
181CTCGTGACCACCCTGACCTACGGCGTGCAGTGCTTCAGCCGCTACCCCGACCACATGAAG
241CAGCACGACTTCTTCAAGTCCGCCATGCCCGAAGGCTACGTCCAGGAGCGCACCATCTTC
301TTCAAGGACGACGGCAACTACAAGACCCGCGCCGAGGTGAAGTTCGAGGGCGACACCCTG
361GTGAACCGCATCGAGCTGAAGGGCATCGACTTCAAGGAGGACGGCAACATCCTGGGGCAC
421AAGCTGGAGTACAACTACAACAGCCACAACGTCTATATCATGGCCGACAAGCAGAAGAAC
481GGCATCAAGGTGAACTTCAAGATCCGCCACAACATCGAGGACGGCAGCGTGCAGCTCGCC
541GACCACTACCAGCAGAACACCCCCATCGGCGACGGCCCCGTGCTGCTGCCCGACAACCAC
Amplicon · 360 bp
CCATCCTGGTCGAGCTGGACGGCGACGTAAACGGCCACAAGTTCAGCGTGTCCGGCGAGGGCGAGGGCGATGCCACCTACGGCAAGCTGACCCTGAAGTTCATCTGCACCACCGGCAAGCTGCCCGTGCCCTGGCCCACCCTCGTGACCACCCTGACCTACGGCGTGCAGTGCTTCAGCCGCTACCCCGACCACATGAAGCAGCACGACTTCTTCAAGTCCGCCATGCCCGAAGGCTACGTCCAGGAGCGCACCATCTTCTTCAAGGACGACGGCAACTACAAGACCCGCGCCGAGGTGAAGTTCGAGGGCGACACCCTGGTGAACCGCATCGAGCTGAAGGGCATCGACTTCAAGGAGG

4 · Thermocycler

Denaturation
95 °C melts the hydrogen bonds — the duplex separates into single strands.
95.0 °C
cycle 1 / 30
95°C72°C53°C
95 °C denature → 53 °C anneal → 72 °C extend · last 6 cycles
Amplicon copies per starting template (log scale)
Ideal copies (2ⁿ)
1.00e+0
after 0 cycles
Realistic copies
1.00e+0
95% efficiency, decaying toward plateau
Exact-length strands
0.00e+0
2ⁿ⁺¹ − 2n − 2 (the rest are long, ragged products)

Doubling every cycle is the theory. In a real tube, primers and dNTPs run out, the polymerase loses activity after ~30 min above 90 °C, and the sheer number of product strands means they re-anneal to each other faster than primers can find them. Amplification flattens into the plateau phase — which is exactly why endpoint PCR cannot tell you how much template you started with, and why qPCR reads the exponential phase instead.

Why PCR runs biotech

It makes DNA countable. A single molecule becomes billions in an afternoon. Diagnostics (every COVID PCR test), forensics, ancient-DNA sequencing and prenatal screening all exist because you can take a vanishing amount of template and make it measurable.

It makes DNA editable. Primers are not just copies — you can hang extra sequence off their 5' ends. Add a restriction site, a His-tag, a promoter or a Gibson overlap and the amplicon comes out pre-loaded with it. Nearly every plasmid ever built passed through a primer someone designed.

And it is where CRISPR gets checked. After an edit, you PCR across the target site and sequence the amplicon to see what the cell actually did. Bad primers here mean a false result — which is why the boring checks above (Tm match, GC clamp, no dimers) are the difference between a clean band and a wasted week.