This is the denovo pipeline from the Sequana project.

If you already have all requirements, install the package with pip:

pip install sequana_denovo --upgrade

You will need third-party tools (spades, prokka, quast, etc.). Install all dependencies at once:

mamba env create -f environment.yml

Scan FastQ files in a directory and set up the pipeline (replace DATAPATH with your input directory):

sequana_denovo --input-directory DATAPATH

To skip Prokka annotation:

sequana_denovo --input-directory DATAPATH --skip-prokka

To tune SPAdes memory (default 64 Gb) and digital normalisation:

sequana_denovo --input-directory DATAPATH --spades-memory 32 --digital-normalisation-max-memory-usage 1e9

This creates a denovo/ directory with the pipeline and configuration file. Execute the pipeline locally:

cd denovo
sh denovo.sh

If you are familiar with Snakemake, you can also run the pipeline directly:

snakemake -s denovo.rules --cores 4 --stats stats.txt

See .sequana/profile/config.yaml to tune Snakemake behaviour (cores, cluster settings, etc.).

With apptainer, initiate the working directory as follows:

sequana_denovo --input-directory DATAPATH --use-apptainer

Images are downloaded in the working directory. To store them in a shared location:

sequana_denovo --input-directory DATAPATH --use-apptainer --apptainer-prefix ~/.sequana/apptainers

Then run as usual:

cd denovo
sh denovo.sh

This pipeline requires the following executables (install via bioconda/conda):

• spades or unicycler — de-novo assembler (--assembler option)
• khmer — digital normalisation (normalize-by-median.py, filter-abund.py, etc.)
• quast — assembly quality assessment
• prokka — genome annotation (optional, --skip-prokka)
• busco — assembly completeness assessment (optional)
• checkm-genome — genome completeness and contamination (optional)
• bwa + sambamba — read mapping back to assembly
• freebayes — variant calling
• samtools — BAM/SAM processing
• seqkit — contig filtering by length
• blast — taxonomic identification of contigs (optional)
• multiqc — aggregated HTML report
• graphviz — pipeline DAG image

This Snakemake pipeline assembles bacterial (or other small) genomes from short Illumina reads.

Digital normalisation (khmer): optionally reduces sequencing depth to a target coverage level, discarding redundant reads. This lowers memory usage and speeds up assembly without significantly impacting quality.

Assembly: SPAdes (default) or Unicycler. SPAdes uses multiple k-mer sizes and is recommended for most bacterial genomes. Unicycler is designed for hybrid or circular assemblies.

Quality assessment (QUAST): reports assembly statistics (N50, # contigs, total length, GC%, coverage depth) with an interactive Icarus contig browser.

Annotation (Prokka): rapid prokaryotic genome annotation producing GFF, GenBank, and other standard formats.

Coverage analysis (sequana_coverage): reads are mapped back to the assembly with BWA, duplicates flagged with Sambamba, and per-contig coverage profiles computed and visualised.

Variant calling (Freebayes): detects SNPs and small indels between the assembled consensus and the mapped reads.

Completeness (BUSCO / CheckM): optionally assess assembly completeness against conserved single-copy orthologs (BUSCO) or lineage-specific marker genes (CheckM).

Taxonomic identification (BLAST): optionally BLASTs the top contigs against the nt database to identify their taxonomy.

A summary HTML report (summary.html) with per-sample assembly statistics and embedded coverage plots is generated at the end of the run, alongside a MultiQC report.

See the latest documented configuration file for all available parameters.

To contribute to this project, please take a look at the Contributing Guidelines first. Please note that this project is released with a Code of Conduct. By contributing to this project, you agree to abide by its terms.