Output

All results are written to --outdir (default: results). The directory structure is:

results/
├── antismash/
│   ├── <sample>/           # one directory per input genome
│   │   ├── <sample>.json   # antiSMASH JSON (parsed by tabulation scripts)
│   │   ├── <sample>.gbk    # annotated GenBank output
│   │   └── ...
│   └── ...
├── gecco/
│   ├── <sample>/           # GECCO output directory per genome
│   │   ├── *.clusters.tsv
│   │   ├── *.features.tsv
│   │   └── *.region*.gbk   # generated when BiG-SCAPE or BiG-SLiCE is enabled
│   └── ...
├── deepbgc/
│   ├── <sample>/           # deepBGC output directory per genome
│   │   ├── *.bgc.tsv
│   │   ├── *.full.gbk
│   │   └── *.region*.gbk   # converted to antiSMASH-like region files
│   └── ...
├── combgc/
│   ├── <sample>/
│   │   ├── combgc_summary.tsv
│   │   └── combined_regions/  # representative non-redundant region GBKs used for clustering
│   └── ...
├── bigscape/               # only when --bigscape_run
│   └── output_files/
│       ├── *.network       # GCF network files (one per class + mix)
│       ├── *.tsv           # cluster annotation tables
│       └── ...
├── bigslice/               # only when --bigslice_run
│   ├── result/             # SQLite database and analysis outputs
│   └── ...
├── summary/
│   ├── all_regions.tsv     # per-BGC-region table
│   └── region_counts.tsv   # per-genome BGC count table
└── pipeline_info/
    ├── execution_timeline_<timestamp>.html
    ├── execution_report_<timestamp>.html
    ├── execution_trace_<timestamp>.txt
    └── pipeline_dag_<timestamp>.html

---

summary/all_regions.tsv

One row per biosynthetic region across all input genomes. Produced by TABULATE_REGIONS.

Column	Description
file	Input genome identifier (stem of the antiSMASH JSON filename)
record_id	Sequence record name (e.g. NCBI accession)
region	Region number within the record
start	Start coordinate (bp, 0-based)
end	End coordinate (bp)
contig_edge	True if the region extends to the edge of the contig (potentially truncated)
product	BGC product class(es), separated by / for hybrids
knownclusterblast_hit	Top KnownClusterBlast hit name (only present when --antismash_cb_knownclusters)
knownclusterblast_accession	MIBiG accession of the top hit
knownclusterblast_similarity	Similarity category: high (>75 %), medium (>50 %), or low (>15 %)
record_desc	Full sequence record description

!!! note The three knownclusterblast_* columns are only included when --antismash_cb_knownclusters is set.

---

summary/region_counts.tsv

One row per genome assembly (or contig when --count_per_contig). Produced by COUNT_REGIONS.

Column	Description
record	Genome assembly name (or genome\|contig when --count_per_contig)
total_count	Total number of BGC regions regardless of class
<bgc_type>	Count of regions of that product type (one column per distinct type observed across all genomes)
hybrid	Count of multi-product regions (absent when --split_hybrids)
description	Sequence description; includes [N total records] suffix for multi-contig assemblies

---

antismash/<sample>/

Raw antiSMASH output for each genome. The key file is <sample>.json, which is read by the tabulation scripts. Other files (HTML, GenBank, SVG plots) are present depending on the antiSMASH flags used.

---

gecco/<sample>/

Raw GECCO output for each genome. When --bigscape_run or --bigslice_run is enabled, clystere also runs gecco convert gbk --format bigslice and publishes GECCO region GenBank files compatible with clustering tools.

---

deepbgc/<sample>/

Raw deepBGC output for each genome (*.bgc.tsv and *.full.gbk). clystere also generates antiSMASH-like *.regionNNN.gbk files using BiG-SLiCE's conversion script to ensure compatibility with BiG-SCAPE/BiG-SLiCE.

---

combgc/<sample>/

Unified per-sample BGC selection generated from antiSMASH, GECCO, and deepBGC predictions. The combined_regions/ directory contains representative region GenBank files used as clustering input.

---

bigscape/output_files/

Standard BiG-SCAPE output. The .network files are tab-separated edge lists suitable for import into Cytoscape or Python networkx. One network is generated per BGC class plus a mix network when --bigscape_mix (default).

---

bigslice/

Standard BiG-SLiCE output generated from the unified combgc/*/combined_regions/ folders. The directory includes the processed analysis results and SQLite-backed data used for downstream inspection.

---

pipeline_info/

Nextflow execution metadata generated per run:

File	Description
execution_timeline_*.html	Gantt chart of process execution times
execution_report_*.html	Resource usage report (CPU, memory, I/O per process)
execution_trace_*.txt	Raw per-task resource trace (parseable TSV)
pipeline_dag_*.html	Directed acyclic graph of the pipeline