From One Variant to the Oligos That Build and Read It
6 min read · Updated October 30, 2026
A variant arrives as a coordinate and an allele. What a bench needs is a tube of oligos: the primers that install the change, and the primers that read it back afterwards.
Those are two designs on one piece of sequence, and putting them in one table exposes a few things that are easy to get wrong when they are done separately.
The reference allele is checked before anything is designed
The first thing done with a coordinate is to look at what is actually there. Give position 300 of a 600 bp reference as a T when it reads G, and the design is refused with the bases found and their context quoted: the sequence around it reads CGCTTTGGCC[G]CCGCCCAGTC.
This matters because a wrong coordinate does not fail loudly on its own. It is right for a different isoform, assembly or strand, and it produces a perfectly valid-looking primer set for the wrong base — a set you would order, receive and use before anything went visibly wrong.
A VCF-style anchored pair is normalised first: GC to AC at position 300 is trimmed to G to A at 300, so the design is the substitution it actually is rather than a two-base replacement.
Five oligos for one substitution
For a single-base substitution with enough flanking sequence, the order table holds five rows: a forward and reverse mutagenic primer that install the change, two allele-specific forward primers carrying the standard universal tails — one for each allele — and one common reverse primer.
In the worked example, the mutagenic pair is 31 nt each with a Tm of 81.6 °C and a recommended annealing temperature of about 58 °C, in the overlapping QuikChange style. The allele-specific primers come out at 39 nt, of which 21 are the universal tail, with a 72 bp genotyping amplicon.
Every row carries its sequence as ordered, its length, its Tm, and — crucially — the conditions that Tm was computed under.
The two Tm columns are not on the same scale
The mutagenesis rows report a nearest-neighbour Tm of the whole primer as ordered, salt-corrected for 250 nM oligo, 50 mM Na+, 1.5 mM Mg2+ and 0.2 mM dNTP. The two allele-specific rows report the Tm of the 18 nt core alone, at a 1 M Na+ reference state with no salt correction, because the 21 nt universal tail is not templated in the first round and must not be counted; the common reverse primer on the same basis reports its whole length.
For the same oligo those two conventions differ by several degrees, so the column cannot be read down. 81.6 °C and 72.5 °C in one table do not mean the first primer binds more tightly; they mean two different questions were asked.
Reporting one Tm for an oligo and a differently-computed Tm for the next is how a plan stops being trustworthy, which is why each row names its basis instead of quietly harmonising them.
An indel gets mutagenesis primers and no genotyping pair
Allele-specific genotyping discriminates two alleles by a 3'-terminal base that matches one and mismatches the other. Only a single-base substitution has such a base, so for an insertion, deletion or multi-base replacement there is nothing to put at the primer's 3' end.
The right output is therefore a named omission rather than an empty list. A six-base insertion comes back with the mutagenesis half designed, the genotyping half declined, and the reason stated: the alleles differ by 6 bp, so a flanking PCR sized on a gel or a fragment analyser genotypes it directly, as does amplicon sequencing.
An empty genotyping section would have read as a tool that failed. A named omission with the alternative readout is the answer.
What is not checked
Specificity. Every primer is designed against the reference you supplied and nothing else, so an oligo that also binds three other places in your plasmid, genome or transcriptome passes everything here. There is no BLAST and no off-target search of any kind.
Secondary structure and cross-dimerisation are likewise outside this step. Both are reasons to put the finished table through an oligo analyser before ordering, and neither is a reason to distrust the design — they are a different question about the same sequences.
And no check establishes that the sequence you supplied is the molecule in your freezer. A coordinate that happens to match at the wrong position still passes.
Frequently asked questions
What does a variant-to-oligos plan actually give me?
For a single-base substitution: two mutagenic primers that install the change, two allele-specific primers carrying the universal genotyping tails, and one common reverse primer — five rows, each with its sequence, length, Tm and the conditions that Tm was computed under, plus the edited sequence and a recommended annealing temperature.
Why can I not compare the Tm values down the table?
Because two conventions are in use and each row says which. Mutagenesis rows are salt-corrected nearest-neighbour Tms of the whole primer; genotyping rows are the untailed allele-specific core at a 1 M Na+ reference state, which runs several degrees higher for the same oligo. Put them on one basis with an oligo analyser if you need to compare.
My variant is an insertion. Why are there no genotyping primers?
Allele-specific PCR works by a 3'-terminal base matching one allele and mismatching the other, which only exists for a single-base substitution. For an indel the design is declined by name, with the readout that does work instead — a sized flanking PCR when the alleles differ by enough base pairs, or amplicon sequencing.
What goes wrong when a variant's coordinate is slightly off?
A coordinate that is right for something else. It refuses when the base you declared is not the base at that position, quoting what is actually there and the surrounding sequence — the usual cause being a different isoform, assembly or strand. Without that check the output is a valid primer set for the wrong base.
Related tools
Turn one variant description into one oligo order table: the primers that build it and the primers that genotype it.
Design SDM primers from a nucleotide or amino-acid change, QuikChange or Q5 style.
Design KASP/ARMS SNP-genotyping primers — two allele-specific forward primers with FAM/HEX tails and an engineered ARMS secondary mismatch, plus a common reverse primer.
Nearest-neighbor Tm, ΔG, hairpins and primer dimers for any oligo or primer pair.
Convert HGVS between c., g. and p.: a c. variant to genomic coordinates and its protein effect, or a genomic position back onto its MANE transcript — via a live Ensembl exon map.