Clone the repository and install StrainMake:
git clone https://github.com/UMMISCO/strainmake.git
cd strainmake/
pip install -e .
strainmake --versionEnsure to have at least Snakemake and Conda installed.
You can use the Docker image and run everything via a Docker container (strainmake is the entrypoint):
docker pull bapt931894/strainmake:latest
# show CLI help
docker run --rm bapt931894/strainmake --help
# run the pipeline (example)
docker run --rm bapt931894/strainmake run --cores 16If you need raw Snakemake in the container, override the entrypoint:
docker run --rm --entrypoint /opt/conda/envs/snakemake8.24.1/bin/snakemake bapt931894/strainmake -hNote that you should mount volumes for keeping the generated data:
docker run bapt931894/strainmake \
-v /where/to/keep/results:/opt/strainmake/results \
-v /where/to/keep/logs:/opt/strainmake/logs \
-v /where/to/keep/benchmarks:/opt/strainmake/benchmarks \
...Configuration, paths, etc., should also be consistent with the container file context.
A step by step example of use is available on the wiki.
| Tool | First release | Conda available? | Link | Implemented? |
|---|---|---|---|---|
| fastp | 2018 | Yes | https://github.com/OpenGene/fastp | Yes |
| fastQC | 2010 | Yes | http://www.bioinformatics.babraham.ac.uk/projects/fastqc/ | Yes |
| Tool | First release | Conda available? | Link | Implemented? |
|---|---|---|---|---|
| bowtie2 | 2012 | Yes | https://github.com/BenLangmead/bowtie2 | Yes |
Human assembly for mapping:
- GRCh38. Link
| Tool | First release | Conda available? | Link | Implemented? |
|---|---|---|---|---|
| MEGAHIT | 2015 | Yes | https://github.com/voutcn/megahit | Yes |
| (Meta)SPAdes | 2017 | Yes | https://github.com/ablab/spades | Yes |
| (Meta)Flye | 2020 | Yes | https://github.com/mikolmogorov/Flye | Yes |
| HyLight | 2024 | Yes | https://github.com/LuoGroup2023/HyLight | Yes |
| Tool | First release | Conda available? | Link | Implemented? |
|---|---|---|---|---|
| QUAST | 2013 | Yes | https://github.com/ablab/quast | Yes |
| Tool | First release | Conda available? | Link | Implemented? |
|---|---|---|---|---|
| Prodigal | 2010 | Yes | https://github.com/hyattpd/Prodigal | Yes |
| CD-HIT | 2012 | Yes | https://github.com/weizhongli/cdhit | Yes |
| Tool | First release | Conda available? | Link | Implemented? |
|---|---|---|---|---|
| MetaBAT 2 | 2019 | Yes | https://bitbucket.org/berkeleylab/metabat/src/master/ | Yes |
| SemiBin2 | 2023 | Yes | https://github.com/BigDataBiology/SemiBin | Yes |
| VAMB | 2019 | Yes | https://github.com/RasmussenLab/vamb | Yes |
| Tool | First release | Conda available? | Link | Implemented? |
|---|---|---|---|---|
| CheckM2 | 2023 | Yes | https://github.com/chklovski/CheckM2 | Yes |
| Tool | First release | Conda available? | Link | Implemented? |
|---|---|---|---|---|
| Binette | 2024 | Yes | https://github.com/genotoul-bioinfo/Binette | Yes |
| Tool | First release | Conda available? | Link | Implemented? |
|---|---|---|---|---|
| GTDB-Tk | 2022 | Yes | https://github.com/Ecogenomics/GTDBTk | Yes |
| Tool | First release | Conda available? | Link | Implemented? |
|---|---|---|---|---|
| dRep | 2017 | Yes | https://github.com/MrOlm/drep | Yes |
| Tool | First release | Conda available? | Link | Implemented? |
|---|---|---|---|---|
| Prodigal | 2010 | Yes | https://github.com/hyattpd/Prodigal | Yes |
| Tool | First release | Conda available? | Link | Implemented? |
|---|---|---|---|---|
| CheckM | 2015 | Yes | https://github.com/Ecogenomics/CheckM | Yes |
| Tool | First release | Conda available? | Link | Implemented? |
|---|---|---|---|---|
| CarveMe | 2018 | Yes | https://github.com/cdanielmachado/carveme | Yes |
| Tool | First release | Conda available? | Link | Implemented? |
|---|---|---|---|---|
| MetaPhlAn | 2023 | Yes | https://github.com/biobakery/MetaPhlAn | Yes |
| Meteor2 | 2024 | Yes | https://github.com/metagenopolis/meteor | Yes |
| StrainScan | 2023 | Yes | https://github.com/liaoherui/StrainScan | Yes |
| Tool | First release | Conda available? | Link | Implemented? |
|---|---|---|---|---|
| inStrain | 2021 | Yes | https://github.com/MrOlm/inStrain | Yes (for SR only) |
| Floria | 2024 | Yes | https://github.com/bluenote-1577/floria | Yes (for SR only) |
You can find in workflow/scripts/other_scripts scripts made for processing some results produced by this pipeline.
skani_analysis.py performs bins pairwise comparison using Skani (https://doi.org/10.1038/s41592-023-02018-3). It can also produce a Venn diagram for results derived from dereplicated bins.
usage: skani_analysis.py compare [-h] --bins {refined,dereplicated} --tmp TMP --output_file OUTPUT_FILE --tsv_output TSV_OUTPUT --ani_threshold ANI_THRESHOLD --json_output JSON_OUTPUT --venn_diagram
VENN_DIAGRAM --cpu CPU
options:
-h, --help show this help message and exit
--bins {refined,dereplicated}
Type of bins to analyze
--tmp TMP Temporary directory for intermediate files
--output_file OUTPUT_FILE
File to save the output results (Skani matrix)
--tsv_output TSV_OUTPUT
File to save the Skani matrix in TSV format
--ani_threshold ANI_THRESHOLD
Minimal ANI to consider two bins as the same
--json_output JSON_OUTPUT
File to save the bins similarity results according to assembly methods (JSON)
--venn_diagram VENN_DIAGRAM
Where to save the Venn diagram
--cpu CPU Number of CPU cores to use
We can then check bins found from one assembly method only.
usage: skani_analysis.py check [-h] --json_results JSON_RESULTS --tsv_output TSV_OUTPUT --assembly {unique,megahit,metaflye,metaspades,hybridspades}
options:
-h, --help show this help message and exit
--json_results JSON_RESULTS
Path to the JSON produced using "skani_analysis.py compare"
--tsv_output TSV_OUTPUT
File to save the results in TSV format
--assembly {unique,megahit,metaflye,metaspades,hybridspades}
Choose 'unique' to get a list of bins that were not found from at least a second assembly method, at the given ANI threshold you used with the 'compare' subcommand. Chose any other possible assembly method to
get a list of bins recovered from the given assembly (it won't return the redundant bins coming from other asssemblies)
calculate_binned_contigs.py allows to compute the binned rate of contigs, i.e. the percentage of contigs from an assembly that is found in at least one bin at the end.
The script can do it for each sample and its generated contigs for a given assembly method.
usage: calculate_binned_contigs.py [-h] --assembler {megahit,metaflye,hybridspades,metaspades} --results-dir RESULTS_DIR --type {binette,dereplicated_and_filtered} --tsv_output_binned_contigs TSV_OUTPUT_BINNED_CONTIGS
--tsv_output_binned_rate TSV_OUTPUT_BINNED_RATE
Count assembly contigs assigned to a bin.
options:
-h, --help show this help message and exit
--assembler {megahit,metaflye,hybridspades,metaspades}
The assembly we should use
--results-dir RESULTS_DIR
Folder storing the pipeline results. Typically named 'results': /path/to/pipeline/results
--type {binette,dereplicated_and_filtered}
Type of bins
--tsv_output_binned_contigs TSV_OUTPUT_BINNED_CONTIGS
File to save the list of contigs and their number of assignation in bins in TSV format
--tsv_output_binned_rate TSV_OUTPUT_BINNED_RATE
File to save the binning rate of contigs in TSV format
If your sequencing reads have already been preprocessed, you can use strainmake prepare import-preprocessed to set up the results directory so that the pipeline starts directly from the assembly step, using your preprocessed FASTQ files.
To do this, provide a TSV file formatted like config_data.tsv. The script will create symbolic links in the results folder that point to your preprocessed FASTQ files, saving storage space by avoiding unnecessary duplication.
This approach ensures that the pipeline can use your preprocessed data without needing to process the FASTQ files again.
You can make use of strainmake init to build a YAML configuration.
Usage: strainmake init [OPTIONS]
Interactively generate a config.yaml for the StrainMake pipeline.
Walks through each pipeline section (preprocessing, assembly, …) and lets
you accept defaults, edit values, or skip a section entirely.
╭─ Options ───────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────╮
│ * --samples -s PATH Path to the sample metadata TSV (columns: sample, sample_id, type). [required] │
│ --lr-format TEXT Format of long-read sequences: 'fastq' or 'fasta'. [default: fastq] │
│ --output -o PATH Path to write the generated configuration YAML. [default: config.yaml] │
│ --help Show this message and exit. │
╰─────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────╯
Then, to generate the Snakefile with relevant data to be generated by the pipeline, based on the YAML, use strainmake build.
Usage: strainmake build [OPTIONS]
Generate a Snakefile from a StrainMake configuration YAML.
By default all sections present in the config are rendered. Use --perform
to restrict to specific pipeline steps, and --post-processing,
--taxonomic-profiling, or --strain-profiling
for finer granularity within those steps.
Examples:
# All steps (config-driven, default):
strainmake build --config config.yaml
# Only assembly + binning:
strainmake build --config config.yaml --perform assembly --perform binning
# Post-processing with a subset of tools:
strainmake build --config config.yaml \
--perform assembly --perform binning --perform postprocessing \
--post-processing gtdbtk --post-processing carveme
# Taxonomic profiling with Meteor only:
strainmake build --config config.yaml \
--perform taxo_profiling --taxonomic-profiling meteor
# Strain profiling with only inStrain:
strainmake build --config config.yaml \
--perform assembly --perform binning --perform postprocessing \
--perform strain_profiling --strain-profiling instrain
╭─ Options ───────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────╮
│ * --config -c PATH Path to the StrainMake configuration YAML (generated by `strainmake init`). [required] │
│ --output -o PATH Path to write the generated Snakefile. [default: Snakefile] │
│ --perform [preprocessing|assembly|binning|postprocessing|taxo_profiling|strain_profiling] Pipeline step(s) to include. Pass once per step, e.g. `--perform assembly --perform binning`. If omitted, all steps present in the │
│ config are included. │
│ --strain-profiling [instrain|floria] Strain profiling tool(s) to include: instrain, floria. Pass once per tool. Automatically activates 'strain_profiling' even without an │
│ explicit --perform flag. If omitted, all tools are included. │
│ --post-processing [gtdbtk|coverage|carveme|bakta] Post-processing tool(s) to include: gtdbtk, coverage (checkm1), carveme, bakta. Pass once per tool. Automatically activates │
│ 'postprocessing' even without an explicit --perform flag. If omitted, all tools are included. │
│ --taxonomic-profiling [meteor|metaphlan|strainscan] Taxonomic profiling tool(s) to include: meteor, metaphlan, strainscan. Pass once per tool. Automatically activates 'taxo_profiling' even │
│ without an explicit --perform flag. If omitted, all tools are included. │
│ --help Show this message and exit. │
╰─────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────╯
Use strainmake report.
Usage: strainmake report [OPTIONS]
Generate MultiQC report(s) from StrainMake pipeline results.
Collects QC artefacts from preprocessing, assembly, binning and annotation
steps, then runs MultiQC. One report is produced per assembler when
multiple assemblers were used.
╭─ Options ───────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────╮
│ * --results -r DIRECTORY Directory containing StrainMake pipeline results (the `results/` folder). [required] │
│ * --logs -l DIRECTORY Directory containing pipeline logs (the `logs/` folder). [required] │
│ * --output -o PATH Directory where the MultiQC report(s) will be written. [required] │
│ --ani INTEGER ANI threshold used for MAG dereplication (must match the pipeline run). [default: 95] │
│ --multiqc-config PATH Path to the MultiQC YAML configuration file. [default: /path/to/strainmake/workflow/scripts/multiqc_results/multiqc_config.yaml] │
│ --dry-run -n Print the MultiQC commands without executing them. │
│ --help Show this message and exit. │
╰─────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────╯