v0.2.0
Variant Calling snakemake pipeline used at DeMoraes Lab.
It downloads a bacterial reference genome, and uses bowtie2, samtools and varscan to call variants.
Edit config/config.yaml and then run.
snakemake --use-conda --cores all
Specify configuration directly on the command line.
snakemake --cores all --use-conda --config\
reference_id=GCA_013166975.1\
input=input.tsv\
results=results/
snakemake --report report.zip
The input is a tsv file with no headers and the following columns:
- Sample name
- Forward reads paths
- Reverse reads paths
A script is provided to generate this file, to run it.
scripts/generate_input.py --mark1 _R1 --mark2 _R2 tests/data/
The generated file can then be manually edited, for further customization.
The main output is a tsv of the variants, one file per sample.
- An Anaconda Distribution
- I recommended using miniforge
- This README uses mamba, but substitute by conda if appropiatly.
git clone 'https://github.com/elbecerrasoto/variants'
cd variants
mamba env create
mamba activate variants
To test the installation.
make test
The pipeline uses the snakemake conventions,
so you can edit the config file at config/config.yaml
,
and then run:
snakemake --cores all --use-conda
- Use "all" for bowtie cores
- Update readme
- Simplify config
- Simplify input files
- Log generation
- Script for suggesting input
- Workflow catalog standards
bowtie2 uses libcrypt.so.1
you can install it on manjaro like this:
sudo pacman -S core/libxcrypt-compat