Overview

This script automates running Compleasm on multiple genome assemblies (FASTA format) to assess genome completeness using BUSCO lineages. It supports manual lineage selection and auto-detection (via SEPP). It also generates a single summary table with results for all input genomes.

About Compleasm

Compleasm is a tool designed to assess the completeness of genome assemblies based BUSCO lineages. Developed by Neng Huang and Heng Li, Compleasm offers a faster and more accurate alternative to traditional methods like BUSCO.

Installation

1️⃣ Install Compleasm (via Conda/Mamba)

To install Compleasm with mamba, run:

mamba create -n compleasm_env -c conda-forge -c bioconda compleasm
conda activate compleasm_env

2️⃣ Install SEPP (for --autolineage support)

mamba install -c bioconda sepp

Lineage Detection Modes

Compleasm supports two modes for selecting a lineage:

Mode	How to Use	Requirements	Notes
Manual	-l <lineage>	None beyond Compleasm	E.g., -l enterobacterales
Auto-detection	--autolineage	Requires SEPP (mamba install -c bioconda sepp)	Automatically finds the best-matching lineage

Usage

./run_compleasm.sh -i <input_directory> [-l <lineage> | --autolineage] [-o <output_directory>] [-t <threads>]

Arguments

Argument	Type	Required	Description
-i, --input	Path	Yes	Directory containing genome assemblies in .fasta, .fna or .fa format.
-l, --lineage	String	Conditional*	BUSCO lineage name (e.g., enterobacterales, bacteria, etc.).
--autolineage	Flag	Conditional*	Let Compleasm auto-select the best-matching lineage (SEPP required).
-o, --output	Path	No	Output directory for Compleasm results (default: compleasm_out).
-t, --threads	Integer	No	Number of CPU threads (default: 8).
-h, --help	Flag	No	Show help message and exit.

• You must specify either -l or --autolineage (not both).

Example Commands

Run Compleasm with manual lineage:

./run_compleasm.sh -i ./assemblies -l enterobacterales -o klebsiella_results -t 16

Run Compleasm with auto-detection (SEPP required):

./run_compleasm.sh -i ./assemblies --autolineage -o results -t 16

Output

For each input genome:

• Compleasm outputs results in a subdirectory within the output folder.
• Example:

klebsiella_results/
├── genome1/
│   ├── summary.txt
├── genome2/
│   ├── summary.txt

A single summary file is generated:
✅ compleasm_summary.tsv in the output directory
✅ Contains one header and one row per genome
✅ The first column is the basename of the genome FASTA file:

Sample	Complete (%)	Single-Copy (%)	Duplicated (%)	Fragmented (%)	Missing (%)	Total BUSCOs
genome1	98.2	97.0	1.2	1.0	0.8	440
genome2	95.5	93.8	1.7	2.5	2.0	440

⚠️ Note on Lineage Autodetection

While Compleasm supports automatic lineage detection using the --autolineage option (requires SEPP), this feature may not always select the most appropriate lineage for your specific genomes.

For best results, it is recommended to manually specify the lineage using the -l <lineage> option.
For example, for Salmonella or Klebsiella genomes, use:

-l enterobacterales

You can list available BUSCO lineages by running:

compleasm list

Compleasm Command Reference

For advanced options, here’s the compleasm run --help output

usage: compleasm run [-h] -a ASSEMBLY_PATH -o OUTPUT_DIR [-t THREADS]
                     [-l LINEAGE] [-L LIBRARY_PATH] [--odb ODB]
                     [--specified_contigs SPECIFIED_CONTIGS [SPECIFIED_CONTIGS ...]]
                     [--outs OUTS]
                     [--miniprot_execute_path MINIPROT_EXECUTE_PATH]
                     [--hmmsearch_execute_path HMMSEARCH_EXECUTE_PATH]
                     [--autolineage] [--retrocopy]
                     [--sepp_execute_path SEPP_EXECUTE_PATH]
                     [--min_diff MIN_DIFF] [--min_identity MIN_IDENTITY]
                     [--min_length_percent MIN_LENGTH_PERCENT]
                     [--min_complete MIN_COMPLETE] [--min_rise MIN_RISE]

optional arguments:
  -h, --help            Show this help message and exit.
  -a ASSEMBLY_PATH      Input genome file in FASTA format.
  -o OUTPUT_DIR         The output folder.
  -t THREADS            Number of threads to use.
  -l LINEAGE            Specify the name of the BUSCO lineage to be used.
                        (e.g. eukaryota, primates, saccharomycetes etc.)
  -L LIBRARY_PATH       Folder path to download lineages or already downloaded
                        lineages. If not specified, a folder named
                        "mb_downloads" will be created on the current running
                        path by default to store the downloaded lineage files.
  --odb ODB             OrthoDB version, default: odb12.
  --specified_contigs   Specify the contigs to be evaluated, e.g. chr1 chr2.
  --outs OUTS           Output if score at least FLOAT*bestScore [0.99].
  --miniprot_execute_path Path to miniprot executable.
  --hmmsearch_execute_path Path to hmmsearch executable.
  --autolineage         Automatically search for the best matching lineage.
  --retrocopy           Separate retrocopy genes from duplicated genes.
  --sepp_execute_path   Path to run_sepp.py executable.
  --min_diff MIN_DIFF   Threshold for best vs second-best matching.
  --min_identity MIN_IDENTITY The identity threshold for valid mapping results.
  --min_length_percent  The fraction of protein for valid mapping results.
  --min_complete MIN_COMPLETE The length threshold for complete gene.
  --min_rise MIN_RISE   Minimum length threshold for duplication vs single/frag.

Reference

• Huang, N., & Li, H. (2023). Compleasm: a faster and more accurate reimplementation of BUSCO. Bioinformatics, 39(10), btad595. https://doi.org/10.1093/bioinformatics/btad595