Antimicrobial resistance, read from the sequence.

Where catalogues run out

Reference lookup is good at what it was built for. Against organisms and determinants that have been described, it is fast, reproducible and well understood, and nobody should replace it for that.

It has two limits. The first is coverage. A novel variant, a distant homologue or resistance arising from a combination rather than a single gene returns nothing, and in a clinical report nothing reads as susceptible.

The second limit is harder. A catalogue returns presence, and presence is not resistance. A determinant can be silent, truncated, or sitting on a plasmid the organism is no longer maintaining. Presence and phenotype disagree often enough that presence on its own cannot honestly be reported as a result.

What ours returns

An attribution. The protein doing the work, and the drug class it acts against, including in cases where that protein is absent from the reference catalogue.

For a clinical microbiologist that is a different object from a database hit, because it can be acted on directly.

Why the timing is the argument

Phenotypic susceptibility testing is still the reference method and it is still slow. Culture takes days. Our call arrives while empiric therapy is being chosen, which is when a broad-spectrum agent tends to get used out of caution. Narrowing at that point helps the patient in front of you, and it is also the only route by which a stewardship programme reduces selective pressure across a whole hospital.

Evaluating it

Send us a set of isolates you already have phenotypic results for. Compare our attribution against your AST, on your own organisms, before any commercial conversation. That is a better test than anything we could write here.