name: busco description: > Benchmarking Universal Single-Copy Orthologs for genome completeness. Use for installation, parameter choice, execution, output interpretation, or troubleshooting. license: MIT metadata: version: "1.0.1" metagenomics.category: "analysis-tools" metagenomics.tags: "qc, completeness, busco" metagenomics.upstream: "https://gitlab.com/ezlab/busco" metagenomics.stage: "assessment" metagenomics.maturity: "reviewed" metagenomics.last-reviewed: "2026-10-24" metagenomics.command-status: "genome mode and out_path contract statically reviewed; not executed here"

BUSCO

Upstream: BUSCO

Citation

Manni, M. et al. BUSCO update: novel and streamlined workflows along with broader and deeper phylogenetic coverage for scoring of eukaryotic, prokaryotic, and viral genomes. Mol. Biol. Evol. 38, 4647–4654 (2021). https://doi.org/10.1093/molbev/msab199

See also docs/references.md.

Analytical thinking

Useful for isolates/eukaryotes; for prokaryotic MAGs prefer CheckM2 (+ GUNC). Do not mix BUSCO% with CheckM2% unlabeled.

How to run

Select and freeze a domain/lineage dataset before comparisons. BUSCO's -o is a run name; use --out_path for the parent directory.

set -Eeuo pipefail
id="${1:?usage: run-busco.sh GENOME_ID GENOME_FASTA OUTPUT_PARENT}"
genome="${2:?usage: run-busco.sh GENOME_ID GENOME_FASTA OUTPUT_PARENT}"
out_parent="${3:?usage: run-busco.sh GENOME_ID GENOME_FASTA OUTPUT_PARENT}"
lineage="${BUSCO_LINEAGE:-bacteria_odb10}"
threads="${THREADS:-16}"
[[ "$id" =~ ^[A-Za-z0-9][A-Za-z0-9._-]*$ ]] || exit 2
test -s "$genome" || exit 2
test ! -e "$out_parent/$id" && \
  test ! -e "$out_parent/logs/${id}.stdout.log" && \
  test ! -e "$out_parent/logs/${id}.stderr.log" || {
  echo "use a new BUSCO run name" >&2; exit 2;
}
mkdir -p "$out_parent" "$out_parent/logs"
 
busco -i "$genome" -l "$lineage" -o "$id" --out_path "$out_parent" \
  -m genome -c "$threads" --offline \
  >"$out_parent/logs/${id}.stdout.log" \
  2>"$out_parent/logs/${id}.stderr.log"
find "$out_parent/$id" -type f -name 'short_summary*.txt' -size +0c -print -quit | grep -q .
find "$out_parent/$id" -type f -name 'full_table*.tsv' -size +0c -print -quit | grep -q .

Key parameters

Decision notes

• Stage: assessment (gene-set completeness; complementary to CheckM2 for some lineages)
• Prefer lineage-appropriate BUSCO DB; do not treat BUSCO % as MAG contamination
• Pin software + DB versions in Methods

Validation, recovery, and provenance

Require short/full summaries, missing-BUSCO list, logs and lineage metadata. Inspect complete single-copy, duplicated, fragmented and missing fractions, gene-predictor behavior, assembly mode, lineage appropriateness, and sample failures. For MAGs, duplicated BUSCOs are not a calibrated contamination metric; compare only like-for-like lineage and BUSCO releases.

Use fresh run names after changing input, lineage, predictor, or release. Record BUSCO, predictor and dependency versions, lineage name/date/checksum, genome checksum, mode/options, command, logs, summary/full-table checksums, and any exclusions. Keep BUSCO gene-set completeness distinct from CheckM2 estimates.

Related skills

checkm2 · gunc · mag-qc · tool-selection · metagenomics-workflow