No model for your organism, your chemistry, or your construct
A stock basecaller reads what its vendor sequenced. If your material is not in that set the calls do not fail — they degrade quietly, and you find out from a reviewer. Calibration adapts the decoder to your signal, on reads you already have.
When you need this
The Signal Check tells you, so you do not have to guess. Send up to 5 Gsamples and you get one of three verdicts back. Verdict B is this page: your signal decodes, but not well enough, and a calibration set would close the gap. If you already know your material has no model — a non-model plant or insect, an unusual host, a bacterial isolate, a construct with chemistry nobody publishes against — expect B.
What verdict B says
It says which part of your signal we cannot account for, and what would close it. Not a score. You get a written answer in plain language and the option to stop there, which is the point of splitting feasibility from the calibration itself: you decide with a number in front of you rather than after paying for the whole thing.
Which reads to supply
We tell you in writing, before you sequence anything. Usually a few thousand labelled reads from material whose sequence you already know — a control, a reference strain, a plasmid, anything where the truth does not come from us. The spec says how many, from what, and what would make them unusable.
The three stages
- FeasibilityWe run a learning curve on the labelled reads you already have and send a written spec of what, if anything, is missing. You may stop here, and the fee is credited against the calibration if you do not.
- CalibrationWe adapt the decoder's front end on your labelled reads. The core stays frozen and hash-asserted, so what changed is recorded and what did not change is provable.
- ReuseYou receive a calibration ID. Every later Decode Run or Specified Search Run on that pore and chemistry reuses it at list price, with no second calibration.
What it costs
Feasibility €900, credited in full against the calibration. Calibration €4,900, one time per pore and chemistry, including a decode of your data.
What it is meant for
- An animal isolate or host with a reference, where the method has to stand up in a validation file.
- A plant pathogen, a cultivar, or a GMO event with a reference.
- A bacterial isolate with a known methylome, where the question is 6mA and 4mC motifs rather than the sequence.
- A genome at an extreme of GC content, where a stock model's errors are not evenly spread.
Where calibration stops
Calibration adapts a decoder to a signal it can already partly read. If the alphabet itself is new — a base the decoder has never been shown — there is nothing to adapt from, and the answer is an Alphabet Extension instead. That is verdict C, and the Signal Check tells you which of the two you are in before you spend anything.
If you cannot make the calibration set
We can sequence your control material in our lab as part of the calibration. It is the one part of what we do that is not digital only, and it exists because some labs have the material but not the run.
Start here
The free Signal Check tells you which verdict you are in, and the feasibility stage tells you what a calibration would involve before you commit to one. Non-human data: nothing to sign, just the request form.
Patient specimens and material from identifiable persons are not accepted.