Site avoidance heat map
Draws the last site-avoidance sweep: every record against every recognition site, coloured by how far its count sits below or above the null.
What it does
Opens beside the Site Avoidance panel when a sweep runs, and redraws when another runs. Each row is a record of the corpus and each column a recognition site, so a site avoided across a whole lineage reads as a vermillion column and a record with its own pattern as a row that stands out. Hover a cell for its numbers; click a cell in the row of the record open in the editor to draw that site's copies on the map.
What this computes
Nothing new: every cell is a value the sweep returned.
- log₂ observed-to-expected ratio of each (record, site) test, as the cell's colour: vermillion below the null, blue above it, full colour at the run's own largest |log₂ ratio|
- zero-copy sites, painted at the avoided end, since their ratio has no finite logarithm
- the avoided and enriched calls, a dot in each cell whose q-value passed the false discovery rate the run set
- column order, sites from the most often called avoided to the most often called enriched, then by their mean ratio across records
- untestable cells, left empty: no copy expected, or a site the maximal-order null cannot score
Limitations
- The colour is the ratio, not the evidence: a short record can show a strong colour on a site whose expected count is too small for any call. The dot is the call.
- The scale is relative to the run, so colours are comparable within one sweep, not across sweeps.
RNA-seq
Draws the last RNA-seq run: the volcano, MA plot or dispersion fit of a differential-expression run, the strongest over-represented gene sets, or one gene set's running enrichment score along the ranking.
Protein coverage
Draws each inferred protein's sequence with the identified peptides laid under the residues they cover, marking which peptides it shares with other proteins.