Overview
A curated, evidence-linked knowledge base of 609 fungal biosynthetic gene cluster (BGC) dossiers, each anchored to a MIBiG accession and built from primary literature with full source provenance. Snapshot: 2026-08-21.
Taxonomy Explorer
Chemistry & Biosynthesis
Use the dropdown to switch between the four independent categorical descriptors MIBiG/curation reports for each BGC.
Facet Explorer
609 BGCs are independently categorised along 27 facet axes (biosynthesis, chemistry, ecology, activation conditions, genome context). Pick an axis to see its value distribution and the matching dossiers.
Knowledge Graph
Every RELATIONSHIP row is a curated subject–predicate–object triple (e.g. gene encoded_by protein, cluster produces metabolite). Click a predicate in the legend to isolate it.
Evidence & Provenance
Conflicts & Uncertainty
Disagreements between sources, boundary ambiguities and unresolved causal questions are recorded as first-class CONFLICT rows rather than silently collapsed into a single consensus value — by design.
Dossier Browser
Search and filter all 609 BGC dossiers; click a row to assemble its full cross-table
record — organism, genes, metabolites, experiments, measurements, claims, relationships,
conflicts and sources, joined on dossier_id.
| Dossier | MIBiG | Organism | Product | Year | Genes | Metab. | Conflicts |
|---|
Select a dossier to view its full record.
Data Dictionary
About this resource
Scope — 609 fungal biosynthetic gene-cluster dossiers, each linked to a MIBiG accession.
Data organization — Separate tables contain BGCs, organisms, strains, facets, genes, metabolites, experiments, experimental terms, measurements, observations, claims, relationships, conflicts, sources and evidence links.
Source traceability — SOURCE and EVIDENCE_LINK contain literature and database provenance.
Raw data — RAW_ROW retains non-empty biological source rows.
Normalization — TERM contains raw and canonical term labels; NORMALIZATION_QUEUE contains mappings that remain unresolved.
Missing or uncertain values — Explicit missingness and uncertainty states are retained where present.
Snapshot — Static shareable copy created 21 August 2026. The working normalization file remains separate.
Schema (23 tables)
| table | unit | purpose |
|---|---|---|
| BGC | one row per dossier/BGC | stable identifiers and minimally interpreted core descriptors |
| ORGANISM | one row per taxon-resolved organism key | taxonomy entity separated from BGC |
| STRAIN | one row per exact strain label within organism context | strain entity for experiment linkage |
| BGC_FACET | one BGC × independent categorization axis × value | multi-axis categorization without forcing a single BGC class |
| GENE | one dossier-local gene record | gene architecture/function evidence |
| METABOLITE | one dossier-local metabolite record | product/intermediate evidence |
| EXPERIMENT | one experimental event | event envelope; conditions/actions linked separately |
| TERM | one exact typed raw concept | vocabulary layer for conditions, interventions, assays and time |
| EXPERIMENT_LINK | one experiment × typed term link | hyperedge membership; separates action/state/assay/context components |
| MEASUREMENT | one reported measurement | quantitative/text measured value with explicit subject |
| OBSERVATION | one outcome or measurement observation | observation layer separated from actions/state |
| CLAIM | one claim | reported/synthesized claims separate from measured observations |
| RELATIONSHIP | one explicit relation | graph-ready relation; free-text relations remain unresolved until normalized |
| CONFLICT | one conflict/uncertainty | preserve disagreement rather than collapse it |
| SOURCE | one dossier-local source | literature/database provenance |
| EVIDENCE_LINK | one record × source link | many-to-many provenance |
| ASSERTION | one scalar field/value assertion | atomic long-form fact ledger across schema generations |
| COMPACT_FACT | one compact legacy row | structured salvage layer for older dossier generations; evaluative label column excluded |
| NORMALIZATION_QUEUE | one unresolved row/reference | explicit backlog; no forced interpretation |
| PROVENANCE_EXCLUDED | one excluded evaluative/administrative location | track what was excluded without carrying the bias-prone value |
| RAW_ROW | one nonempty biological source row | lossless audit layer for not-yet-normalized information |
Feature policy (data families)
| Data families | Contents |
|---|---|
| Identifiers and administration | Dossier IDs, accessions, source metadata and record identifiers |
| BGC facets | Independent categorical descriptors attached to a BGC |
| Experimental terms | Conditions, interventions, media, developmental states, time, sample matrices and analytical methods |
| Measurements and observations | Quantitative measurements and qualitative experimental outcomes |
| Sources and evidence links | Literature/database records and links from scientific records to supporting sources |
| Missingness and uncertainty | Explicit recorded states such as NR, NF, NA, UNCERTAIN and CONFLICTING |
View groups
| Table group | Main tables |
|---|---|
| BGC identity and context | BGC, ORGANISM, STRAIN, BGC_FACET |
| Genes and products | GENE, METABOLITE |
| Experiments and conditions | EXPERIMENT, TERM, EXPERIMENT_LINK |
| Experimental results | MEASUREMENT, OBSERVATION |
| Interpretive records | CLAIM, RELATIONSHIP, CONFLICT |
| Sources and traceability | SOURCE, EVIDENCE_LINK, ASSERTION |
| Legacy / unresolved / raw | COMPACT_FACT, NORMALIZATION_QUEUE, PROVENANCE_EXCLUDED, RAW_ROW |