Then replace the example with your own sequence and mutations.
Paste the coding sequence of your gene, from the A of the start codon, in frame. Raw sequence or FASTA both work; FASTA header lines beginning with > are stripped automatically, whether you paste them or load them from file. All non-ACGT characters (whitespace, line numbers, ambiguity codes) are discarded before translation.
📂 Load FASTA accepts .fasta, .fa, and .txt. The nucleotide count appears next to the buttons once a sequence is loaded — check it is a multiple of 3 and matches your expected ORF length.
⚠️ Paste the gene, not the whole plasmid. Numbering and translation both assume the sequence starts at codon 1 (or at a fixed offset — see Frame shift).
Selects the codon assigned to each substituted amino acid.
| Option | Table | Source |
|---|---|---|
| E. coli | E. coli K-12 most frequent codons | Hershberg & Petrov 2008 |
| S. cer. | S. cerevisiae most frequent codons | Sharp & Li 1987 |
| P. pas. | P. pastoris (K. phaffii) most frequent codons | Xu et al. 2021 |
One preferred codon per amino acid is used; the tool does not sample alternative synonymous codons.
Standard one-letter protein nomenclature: [WT residue][position][new residue], e.g. D110A. Separate entries with commas, spaces, or newlines — all three are accepted interchangeably.
The WT residue you type is verified against the translated sequence. A mismatch raises an error rather than designing a wrong primer, so it doubles as a frame check.
Degenerate codons for saturation mutagenesis use two reserved letters in the "new residue" position:
| Code | Codon | Coverage |
|---|---|---|
X |
NNK | All 20 amino acids, 32 codons, one stop (TAG) |
Z |
NDT | 12 amino acids, 12 codons, no stops, no redundancy |
So G21X orders an NNK library at position 21; G21Z orders the balanced NDT subset. Both can be listed alongside defined mutants at the same site — see Shared primers.
Determines which primer carries the mutation. This is the single most consequential choice for cost.
| Mode | Behaviour | Use when |
|---|---|---|
| ⇄ Both | Both primers carry the mutant codon; symmetric overlap design | Single defined mutants; most robust amplification |
| → Fwd only | FWD carries the mutation; REV is pure wild-type | Several variants at one site — the WT reverse primer is shared |
| ← Rev only | REV carries the mutation; FWD is pure wild-type | As above, mirrored; useful if the downstream flank has poor Tm or a repeat |
In fwd/rev mode the wild-type primer is identical for every variant at that position, so a five-variant site needs six oligos instead of ten.
| Parameter | Default | What it controls |
|---|---|---|
| Frame shift | 0 | Nucleotide offset added to the residue number, for constructs whose numbering does not start at codon 1 — e.g. -9 when the sequence includes a 3-residue N-terminal extension not counted in the published numbering |
| Target Tm (°C) | 60 | Tm target for the priming region of the mutation-carrying primer |
| Min anchor (nt) | 17 | Minimum priming region length, enforced before Tm is checked |
| Tm priming (°C) | Tm − 10 | Tm target for the pure wild-type shared primer in fwd/rev mode. Leave blank to use the default |
Priming regions grow one base at a time from the mutation site until both the minimum anchor length and the Tm target are satisfied, so raising Min anchor and raising Target Tm both lengthen primers — and cost.
| Box | Meaning |
|---|---|
| PROCESSED | Mutations parsed, including failed ones |
| WARNINGS | Designs that succeeded but tripped a quality check |
| ERRORS | Mutations that could not be designed |
| TOTAL COST | Sum of unique primers only, at SGD 0.18/bp |
Within every primer sequence:
For degenerate codons all three bases are uppercase, since every position is variable.
Each card reports, per primer: length, Tm, GC%, and the codon change in the form D(gat)→(gcg)A. The mutated full-length DNA sequence is also stored and exported — use it as the reference for sequencing verification.
| Warning | Threshold | Suggested response |
|---|---|---|
Fwd/Rev self-dimer |
Any 5-nt self-complementary stretch | Usually tolerable; if the PCR fails, shift the anchor by changing Min anchor by ±2 |
Fwd/Rev GC x% |
Outside 40–65% | Raise Target Tm to extend into more balanced flanking sequence |
Fwd/Rev Tm low |
Below 55 °C | Raise Target Tm or Min anchor; check the flank is not AT-rich or repetitive |
Warnings are advisory. A flagged primer is still exported and will often work — treat them as a ranked to-do list if amplification fails.
When a primer sequence has already appeared earlier in the run, it is tagged ♻ Reused from mutation and priced at SGD 0. This is what makes fwd/rev mode economical for site-saturation panels: G21A, G21X, and G21Z in fwd mode all share one reverse primer.
The Order List tab shows only the primers you actually need to buy, named <mutation>_Fwd / <mutation>_Rev; reused ones are excluded.
| Button | File | Contents |
|---|---|---|
| ⬇ Export CSV | primer_design_results.csv |
Full record: both primers, lengths, Tm, GC%, codon change, per-primer cost, warnings, shared-primer notes, mutated DNA sequence, and a total row |
| ⬇ Export Order List | primer_order_list.csv |
Two columns — primer name and sequence — deduplicated, for direct submission to an oligo supplier |
Keep the full CSV with your construct records; it documents exactly what was ordered and why.
| Message | Cause | Fix |
|---|---|---|
Cannot parse position from "..." |
Malformed entry, e.g. Ala110Ala or D-110A |
Use one-letter format: D110A |
Position N out of range (protein length L) |
Residue number beyond the ORF, or a frame problem | Check the sequence is the full CDS and that Frame shift is correct |
Expected D at position 110, found G |
The WT residue does not match the translation at that position | Almost always a numbering offset — set Frame shift. Otherwise confirm you pasted the right gene or the right frame |
Unknown amino acid "B" for this organism |
Letter is not a standard residue, *, X, or Z |
Use a valid one-letter code |
The Expected ... found ... error is the most useful diagnostic in the tool: if it fires on every mutation, your numbering is offset by a constant, and Frame shift will fix all of them at once.
Melting temperature. Nearest-neighbour thermodynamics (SantaLucia 1998) with the Owczarzy et al. 2004 salt correction, evaluated at 50 mM Na⁺ and 250 nM primer. Values are calibrated to agree with Biopython Tm_NN(saltcorr=7).
Priming region growth. Starting from the mutant codon, the region is extended base by base — leftward for the reverse primer, rightward for the forward primer — until both the minimum anchor length and the Tm target are met. For degenerate codons the wild-type sequence is used for the Tm calculation, since nearest-neighbour parameters are undefined for ambiguous bases.
Primer architecture. In all three modes the primer pair shares a 12-nt overlap spanning the mutation site. The amplicon therefore carries a 12-bp terminal direct repeat.
GC content is computed over the whole primer, including the overlap region.
Adapt this section to your lab's standard workflow.
The 12-bp terminal overlap means the linear PCR product is recircularised by overlap-dependent assembly — HiFi/Gibson assembly, In-Fusion, or in vivo recombination in a recombination-competent E. coli strain — rather than by blunt-end ligation, which would duplicate the overlap.
A typical run:
DpnI digestion of the template (37 °C, 1 h).For NNK/NDT libraries, pool sufficient transformants for the target coverage before sequencing a representative sample to confirm codon distribution.
Designed by Simon, Congqiang Zhang, Singapore Institute of Food and Biotechnology Innovation (SIFBI), A*STAR. congqiang_zhang@a-star.edu.sg · Issues and feature requests: GitHub
Released under the MIT License.