Discovery Platform Overview
How the single-screen discovery platform is laid out
The discovery platform works on one screen: you build a query, run it, and triage the results in the same view. This page covers the layout and how its panels connect.
Layout
The screen has a top strip and two working columns: results on the left, detail on the right.
Top strip
Holds the database stat badges (validated, integrated, and predicted BGC counts; genome and metagenome counts), the filter chips that make up your query, the Load Asset control, and the action buttons: Clear query, Generate Report, and Run Query. See Filtering.
Results panel (left)
One panel, three tabs over the same result set:
- BGC roster — a sortable table, one row per iBGC.
- Variables map — a scatter plot of any two metrics you choose.
- UMAP — a 2-D map where similar iBGCs cluster together.
Switching tabs does not change which iBGCs are shown — only how they’re displayed.
Detail panels (right)
Two iBGC detail cards stacked above a protein panel:
- Reference (top) — the iBGC you pin to compare others against.
- Compare (middle) — updates each time you left-click a result.
- Protein Information (bottom, collapsible) — the gene you last clicked, with its domain annotations and sequence.
Both detail cards share the same layout; see iBGC Detail.
The core interactions
The discovery platform is driven by clicks on iBGCs, whether in the roster or on a map:
| Action | Result |
|---|---|
| Left-click an iBGC | Loads it into the Compare panel |
| Right-click → Set as reference iBGC | Pins it to the Reference panel |
| Right-click → Find similar iBGCs | Runs a similarity search around it |
| Right-click → Add to shortlist | Adds it to your shortlist |
| Click a gene (CDS) in a detail card | Loads its protein into the Protein Information panel |
The reference stays pinned while you left-click through other candidates, so you can compare many iBGCs against one anchor.
Nothing loads until you Run Query
The result panels stay empty until you press Run Query. This lets you assemble a complete query — several filters, perhaps an advanced search chip — before anything is fetched. The badges in the top strip always reflect the full catalogue, not your current filters; the match count appears in a banner after you run.
Result caps and sampling
A query can match more iBGCs than the maps can plot at once. When the result set exceeds the display cap, the maps show a representative sample and the banner reports the true total. Narrow the filters to see every match, or work from your shortlist — the report export uses the iBGCs you selected, not the sampled view.