A base nobody has decoded before

What it is

We derive codebook rows for your symbol, adapt the decoder on your calibration data, decode your reads, and deliver the result set, the reconstruction and a written report.

What we start from

Either a molecular structure — a CCD code, a PubChem CID, a MOL2 file or a SMILES string — or a level table you have measured yourself. Plus your own reads containing the symbol, and whatever you know about where in the sequence it sits.

Three milestones, each contracted on its own

• Codebook rowsWe derive the rows for your symbol from the structure or the level table, and say what that derivation does and does not support.
• Calibration and adaptationWe adapt the decoder on your calibration data, with the core frozen and hash-asserted as on every other job.
• Decode and reportWe decode your reads and deliver the result set, the reconstruction and the report.

The honest boundary

A decoder that has never been shown a symbol performs at chance on it. That is the whole reason this is a project rather than a product: the codebook and the adaptation are what move the result off chance, and both need your material. We tell you the held-out result at the end of the second milestone, before you commit to the third.

Calibration or Alphabet Extension?

Calibration is for a signal we can already partly read — your organism or your chemistry is new, but the alphabet is not. An Alphabet Extension is for a symbol that is not in the alphabet at all. The Signal Check separates the two: verdict B is a calibration, verdict C is this.

What it costs

From €9,900, quoted per project. Each milestone is contracted and delivered on its own, so you can stop after any one of them.