Curated reference data are what turn microarray signal into a species name you can stand behind.
A microarray scan is a grid of numbers. The reference library is the curated knowledge that tells the QAID App which numbers mean which organism, how confident that reading is, and what the run should look like when it has gone well.
Each QAID microarray product has its own reference panel: the set of targets it screens for, the probes that represent each target, and the rules that decide when a target counts as identified. The library is that knowledge, held and versioned by QAID.
Panels are built from published DNA sequence data for the relevant taxonomic groups, reviewed against current taxonomy, and checked against reference material of known identity. Candidate probes are designed from that evidence, not adopted from it unexamined.
Public sequence records vary in quality and are sometimes misidentified. Candidate sequences are screened, cross-checked between sources and tested in the laboratory. A probe earns its place on a panel by behaving correctly against both target and non-target material.
Panels and analysis methods carry version numbers, and every result in the QAID App records the versions that produced it. Changes ship as a new version rather than quietly altering how older results would read.
It is easy to find a DNA sequence labelled with a species name. It is harder to be sure the label is right, that the sequence distinguishes that species from its close relatives, and that a probe built from it will still behave correctly in a real, processed sample.
Most of the work behind a QAID panel sits in that gap. Sequences are gathered for the target group and its near neighbours, compared against one another, and checked against current taxonomy, because species names change and records are not always updated. Candidate probes are then tested in the laboratory against material of known identity, including the close relatives most likely to cause a false call.
Several probes stand behind each target on a finished panel. That redundancy is deliberate: a call that rests on agreement between probes is more robust than a call that rests on one spot on a slide.
For the mechanics of how a scan becomes a result, see how QAID works.
Reference data are never finished. New species become relevant, taxonomy is revised, new adulterants appear in a supply chain, and customers ask about materials nobody had raised before. QAID keeps building and re-checking the underlying data, and releases the result as a new panel version for the product it belongs to.
If a target matters to your QA program and you are not sure whether it is covered, ask. That conversation is part of what QAID consulting is for.
The library itself stays proprietary. QAID does not publish probe sequences, panel contents or panel sizes. That design work is what makes a QAID product a product rather than a protocol, and publishing it would tell a supplier exactly what a panel can and cannot see.
What you do get is everything needed to use and defend a result: what was screened for, what was identified, the confidence metrics, the assay-control outcome, the supporting signal data, and the panel and method versions behind them. Validation & quality sets out how those products are shown to work.
Tell us the species, organisms or adulterants that matter in your supply chain, and we’ll tell you which QAID product covers them.