Hit map

What it drawsCopy Anchor Link

The graphic summary of a similarity search — the distribution-of-hits view a sequence-search report opens with. The query is drawn once as a ruler across the full width, and each hit is drawn beneath it as one or more bars, positioned and scaled to the part of the query that hit aligned to.

It answers in one glance what a sorted table of hits cannot: whether your hits cover the whole query or pile onto a single domain, and — the more useful reading — where the query has no coverage at all. A gap in this picture is a region of your sequence that matched nothing, which is often the interesting part.

What opens in itCopy Anchor Link

The result of a similarity search run from the editor. Search results are not a file you open; they are computed or fetched, and this canvas draws what came back.

An alignment of sequences to each other is the alignment canvas. A comparison of two sequences position by position is a dot plot. This canvas is specifically query-anchored.

What this computesCopy Anchor Link

The similarity search itself is run elsewhere; what this canvas computes is the query-anchored geometry of its result, which is a different question from the one the hit table answers.

• Per-hit coverage of the query: where each aligned segment falls on your sequence, in query coordinates, scaled to a ruler drawn at the query's full length. A bar's length is the part of the query involved and says nothing about how much of the database sequence was.
• One bar per high-scoring segment pair, not per hit. A hit made of two separated segments draws two bars on one row — the signature of a repeated domain or an intron-interrupted match — because collapsing them to a single span would erase exactly that.
• Uncovered query regions, which are the negative space the picture is really for: a gap here is a stretch of your sequence that matched nothing, and it is usually the part worth following up.
• Bit score as the shading channel, with its scale stated on the canvas. Bit score rather than raw score because it is normalised for database size, so two searches against differently sized databases shade comparably.
• Strand per segment, marked on the bar, so a hit at the same query position in the opposite orientation is not read as a duplicate of the forward one.
• How many rows the picture can carry honestly. The map draws the strongest hits and stops at a bounded number of rows, because past that point a row is thinner than a line; the full list stays available as a table rather than being drawn illegibly.

Reading itCopy Anchor Link

Every bar is measured against the query, not against the subject. A short bar means the aligned segment covered a small part of your sequence; it says nothing about how much of the database sequence was involved.

One hit can draw several bars. A hit is made of high-scoring segment pairs, and a hit with two separated segments draws two bars on the same row — that pattern is the signature of a repeated domain or an intron-interrupted match, and collapsing it to one bar would erase it.

Shading is bit score, following the convention of the standard report, and the scale is stated on the canvas rather than left to be inferred. Bit score is normalised for database size, which is why it is the shading channel here rather than raw score.

Strand is carried on the bar: a hit on the reverse strand is marked, so a same-position hit in the other orientation is not mistaken for a duplicate.

ControlsCopy Anchor Link

Clicking a bar selects that hit, which drives the detail readout beside the map — the alignment itself, the identities, the expect value. That is the one interaction: the map is a navigation surface into the result, not a place where the result is edited.

The map draws the strongest hits and stops at a bounded number of rows, because past that point the rows are thinner than a line and the picture stops being readable. The full list stays available as a table.

Where it appearsCopy Anchor Link

As a stage tab and inside the search panel that produced it, at panel width. The bars reflow with the box; the query ruler is always the full width, which is what keeps positions comparable between the two.

Performance and limitsCopy Anchor Link

The picture shows position and score. It deliberately does not encode identity percentage or expect value as a visual channel — those are read from the selected hit's detail, because encoding three quantities in one bar produces a drawing that looks precise and cannot be read back.

ScreenshotsCopy Anchor Link

No frame is committed for this canvas yet. What one would show: search or screen hits laid along the sequence they were found on.

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 sequence-search tool that writes it; the alignment canvas, for reading a chosen hit's alignment in full.