Walkthrough: A Worked Example

Start from a known BGC, study its family, and export it

This walkthrough follows one realistic task from start to finish. You have a known, experimentally validated cluster — MIBiG entry BGC0001529 — and you want to find the clusters in the catalogue that are related to it, see which relatives are novel, and learn what the whole family has in common. It is the same path the in-app guided tour takes.

The rationale is the heart of the platform’s report: a single cluster tells you little, but a family of related clusters tells you a lot. Comparing a known BGC against its relatives shows which parts are conserved (the biosynthetic core every member keeps) and which vary (the tailoring that produces different chemistry) — and surfaces the novel relatives worth pursuing.

For the vocabulary used below, read Key Concepts first.

Step 1 — Find the known cluster

Open the Accessions filter and enter the MIBiG assembly accession BGC0001529, then Run Query. The platform resolves it to one iBGC, MGYB-000M9B-01. Right-click the row and choose Set as reference iBGC to pin it to the top detail panel.

The reference card confirms what you started with: a validated cluster (novelty 0 by definition), and — importantly — its gene cluster family, GCF 132.0.0.0.

The reference card for the validated MIBiG cluster. Its GCF chip (132.0.0.0) is the entry point to the rest of the family.

Step 2 — Open up its family

Click the GCF chip (GCF 132.0.0.0) on the reference card. This sets the gene cluster family as a filter. Clear the Accessions chip so you are no longer pinned to the single entry, then Run Query again.

The roster now lists the whole family — the iBGCs whose domain content groups them with your known cluster. Here that is five members: the two validated MIBiG entries (novelty 0) and three candidate clusters from environmental assemblies, with novelty rising to 0.81. Sort by Novelty to bring the most novel relatives to the top.

The five members of GCF 132.0.0.0: the validated reference plus three uncharacterised relatives, the most novel at the top.

This is the payoff of the gene cluster family: one click took you from a single known compound to a set of candidates predicted to make related chemistry — some of which no one has characterised.

Step 3 — Compare the relatives to the known cluster

With BGC0001529 still pinned as the reference, left-click a relative in the roster — start with the most novel, MGYB-000GKN-01. It loads into the Compare card beside the reference, so you can read the two side by side: class, novelty, completeness, and the gene-level maps.

Hover a gene in either map to see its protein domains without leaving the view — a quick way to scan what each cluster encodes.

The validated reference (top) beside a novel relative (bottom). Hovering a gene shows its domains — here, a RiPP leader-peptide domain on the candidate.

For the full picture of a gene, click it. Its protein loads into the Protein Information panel: length, gene caller, every InterPro/Pfam domain, and the copyable sequence. This is where you confirm the enzymatic evidence — for example, that a relative carries the same core biosynthetic domains as the validated cluster but adds tailoring enzymes of its own.

Clicking a gene opens its full protein detail: the InterPro/Pfam domain hits and the sequence.

Step 4 — Shortlist the family

To collect a cluster, open the kebab menu (⋮) on its detail card and choose Add to shortlist.

The kebab menu on a detail card, with “Add to shortlist”.

Repeat for the relatives worth keeping, or use Add all to shortlist in the roster footer to add the whole family at once. The shortlist holds your selection (up to 1,000 iBGCs) in the browser. See Shortlists.

Step 5 — Generate the report

Open the shortlist and click Generate Report. The report analyses your selected clusters as a set — and that is its purpose: to find the common patterns across similar BGCs and flag the outliers.

The family report. The Domain composition panel marks the domains shared by every member (CORE) against the rare ones — the conserved biosynthetic core of the family.

Read it from the top:

  • The iBGC results table lists every member with its scores, class, assembly, and biome.
  • Domain composition is the key panel for this task: it tiers each protein domain as core (present in most or all members), variable, or rare. The core domains are the family’s shared biosynthetic machinery; the rare ones are where individual relatives diverge.
  • GCF and score distributions, completeness, and the biome/taxonomy sunbursts show how broadly the family is spread and how novel its members are.

For this family, the domain composition shows a small set of core domains carried by all five clusters — the conserved heart of the pathway — while the novel relatives add domains the validated entries lack.

Step 6 — Export for further analysis

Click GBKs (zip) in the report header to download a GenBank file per cluster. These are ready for the next step: re-running antiSMASH, aligning the core genes, or designing primers to amplify a candidate from its source organism. See Export Formats for the other formats (TSV, JSON, HTML).

What you’ve done

You started from one characterised compound, gathered its biosynthetic family, compared the relatives against the known reference, and exported the set — with a clear read on which clusters are novel and what machinery they share. The same path works from any starting point: a sequence or chemical search, a taxonomy or biome filter, or your own loaded assembly all lead into the same compare-and-report flow.