Verification

What it drawsCopy Anchor Link

The answer to "is this the molecule I designed?", in one of two forms, because that question is really two questions.

Reads mode grades sequencing reads against a target sequence, one graded row per read, sorted best first. Each row carries a status, an ungapped identity, the fraction of the target the read covers, and counts of mismatches and indels. Selecting a row paints that read's coverage over the target so you can see where it agrees and where it does not.

Construct mode asks the other question: one finished consensus sequence against the design. Strand and origin are resolved first — a circular construct assembled from reads can arrive rotated or reverse-complemented, and comparing before fixing that produces nonsense — then the two molecules are aligned end to end, differences are normalised, and each difference is reported against the feature it lands in.

These are a mode switch rather than two tabs of one form because they take different inputs and answer different questions.

What opens in itCopy Anchor Link

Reads arrive as pasted sequence, as FASTA, as Sanger trace files, or as the per-base coverage table and run summary produced by a whole-plasmid sequencing service. The target is the construct open in the editor unless you paste an expected sequence instead.

A raw trace also opens in the chromatogram canvas, where you look at peak shape; this canvas does not show you peaks, it shows you the verdict.

What this computesCopy Anchor Link

This canvas computes the verdict, and it computes two different ones because "is this the molecule I designed?" is really two questions.

Reads mode, per read, graded and sorted best first:

• Ungapped identity against the target — ungapped on purpose, so a single insertion or deletion does not silently deflate the percentage into looking like widespread disagreement.
• Coverage: the fraction of the target that read spans, which is what separates a read that disagrees from a read that never got there.
• Mismatch and indel counts, reported separately rather than summed into one distance.
• A grade — verified, near, partial or divergent. Partial is not a failure: it usually means one primer read covered part of a long plasmid, which is the normal outcome. Near is the grade to look at, because it means the differences are small enough that a base-calling artefact and a real mutation are both plausible, and only the trace settles it.
• Where each read agrees, painted over the target when you select its row.

Construct mode, one consensus against the design:

• Strand and origin, resolved first. A circular construct assembled from reads can arrive rotated or reverse-complemented, and comparing before fixing that produces nonsense — so rotation and orientation are determined before any difference is counted.
• An end-to-end alignment of the two molecules.
• Normalised differences, so the same change reported at two equivalent positions in a repeat is one difference rather than two.
• The feature each difference lands in, which is what turns a position into a consequence.

Reading itCopy Anchor Link

Four grades, and the middle two are the ones that matter. A read is verified, near, partial, or divergent. Partial is not a failure — it usually means a read covered only part of the target, which is the normal outcome of a single primer read on a long plasmid. Near is the one to look at: it means the read supports the target with differences small enough that a base-calling artefact and a real mutation are both plausible, and the trace is what settles it.

Identity is ungapped, so an insertion or deletion does not silently deflate the percentage — the indel is reported separately as an event, with a position and a length. A read with 100 % ungapped identity and one reported indel is a read with a real indel, not a perfect read.

Coverage is of the target, not of the read. Ninety per cent coverage means ninety per cent of your construct was observed; it says nothing about how much of the read was used.

Where a service summary is loaded, its own quality readings are shown beside the grades rather than folded into them: contamination and mass fractions are evidence about the sample, not about the sequence.

ControlsCopy Anchor Link

Reads can be added, removed and re-run against a different target without losing the others. The report exports as a table. Selecting a row is the main interaction, and it drives the coverage bar over the target.

Nothing here has a threshold you can move to make a read pass. The grade boundaries are fixed because a verification whose pass mark the user sets is not a verification.

Where it appearsCopy Anchor Link

As a stage tab, and docked in the right rail against the open construct — which is where it is usually used, since the target is whatever you have open.

Performance and limitsCopy Anchor Link

Every read is aligned against the full target, so a large batch against a long construct is real work; the panel reports progress rather than appearing to hang.

The honest limit: this canvas grades agreement, not correctness. A read set that agrees perfectly with a target you specified wrongly is graded verified. Construct mode is the guard against that, because it compares against the design rather than against whatever was typed in.

ScreenshotsCopy Anchor Link

No frame is committed for this canvas yet. What one would show: reads placed against a reference with the agreement per position.

The absence is stated rather than left as a blank heading: a picture that does not exist yet is worth knowing about, and a placeholder would be worse than either.

RelatedCopy Anchor Link

The chromatogram canvas, for looking at the peaks behind a near grade; the alignment canvas, for reading several reads against each other rather than each against the target.