Multiplex PCR Panel Design — Primers That Work in One Tube
Choose one primer pair per target so no two cross-dimerise, every amplicon is separable on the gel, and one annealing temperature serves the whole panel.
🌐 Nothing you paste is logged or stored — every tool is also callable via REST & MCP, and in bulk from the batch tools
Designing a multiplex one target at a time is why multiplexes fail. Each target's best pair in isolation is easy; what breaks the reaction is pairwise — a primer for target three cross-dimerising with one for target one, or two amplicons landing forty base pairs apart on the same gel — and neither is visible while you are looking at either target alone. So this searches COMBINATIONS. For every candidate pairing it checks the four primers two targets put in one tube for cross-dimers, judging a dimer that occupies a primer's 3′ end far more strictly than one that pairs internally, because the 3′ end is where extension starts. It checks that the two products would actually resolve on the gel you will pour — a size ratio on the gel's own log scale, inside the resolving window of that agarose percentage, not a rule of thumb about base pairs. And it holds every primer inside one Tm spread, since one annealing temperature has to serve all of them. Then it returns the best feasible combination, or, when there is none, names the target pairs that cannot be multiplexed at all so you know which one to redesign rather than which knob to turn.
How to use the Multiplex PCR Panel Design tool
- 1Add a target per amplicon: its template, optionally the region the product must span, and its own product size band.
- 2Give each target a DIFFERENT size band — that is how a panel is made readable, and leaving them all on one range makes comigration the dominant conflict.
- 3Leave the agarose percentage blank to have the one whose window contains your amplicons chosen and reported, or set it to the gel you will actually run.
- 4Read the panel, or, if none exists, the blocking target pairs — each names the constraint that makes those two mutually exclusive.
Frequently asked questions
Why search combinations instead of scoring each target?
Because the constraints are pairwise and pairwise constraints do not compose. Every target having a good pair, and every pair of targets having some compatible choice, still does not guarantee that one choice per target works for all of them at once — that is the ordinary behaviour of a constraint problem, and the result says so explicitly when it happens. Picking each target's individually best pair optimises the one thing that was never in doubt.
Why does a 3′-end cross-dimer count for more than an internal one?
Because that end is where the polymerase extends from. A dimer that occupies it blocks the reaction the primer exists to start, so it matters at a much weaker binding energy than one that pairs somewhere in the middle and merely takes primer out of circulation. Judging both at the same bar — which the first version of this tool did — rejects about half of all candidate pairings and makes ordinary panels look impossible. Both thresholds are yours to set.
How does it decide two amplicons are distinguishable?
With the same band model the Diagnostic Digest Planner uses, not a rule about base pairs. Migration on agarose is linear in log(size) over the gel's resolving range, so products separate on a size RATIO — and outside that range they do not separate at all, however far apart they look on paper. That second part matters here: a panel of 150-800 bp products judged on a 1% gel has every band below the resolving window, and reads as entirely comigrating. Which is a true statement about a badly chosen gel, so the percentage is picked from your amplicon sizes unless you set it.
Why give each target its own size band?
Because the panel is read by size, so the sizes are a design variable rather than an outcome. Assigning bands — 150 bp for one target, 250 for the next, 400 for the third — is how a readable panel is laid out, and it is what makes the combination search feasible. Leave every target on one range and comigration becomes the binding constraint on almost every pairing.
Does it tell me whether the multiplex will work?
No, and nothing that only reads sequence can. Every constraint checked here is a computed property of the primers — a duplex free energy, a size ratio, a temperature difference. How much each amplicon actually makes when they compete for one polymerase in one tube is not derivable from sequence, so no yield or balance is estimated. What you get is a panel with no detectable reason to fail, which is a different and more honest claim.
What does it not check?
Off-target sites anywhere outside the templates you give it — use the Primer Designer's specificity screen for a genome-wide look. Hairpins and self-dimers inside a single pair, which the primer designer already scores and which show up in each pair's own penalty. And anything about a probe-based multiplex, whose constraints are completely different because the products are not told apart by size at all.
Is the search exhaustive?
At the sizes it accepts, yes, and the result says so. Targets are ordered fewest-candidates-first and branches are pruned as soon as a conflict or the Tm spread rules them out, which keeps the tree small enough to finish; the panel returned is then provably the best under the constraints you set. If the expansion budget is ever reached the result reports that the panel is the best found rather than the best, and lowering the candidates per target restores exhaustiveness.