Tapirs is a reproducible modular workflow for the analysis of DNA metabarcoding data.
Tapirs uses the Snakemake workflow manager and is compartmentalised into several modules, each performing a step of the workflow. Tapirs is designed to be experimental, allowing you to test the effect of different approaches to data analysis. Tapirs is curently v1.0. It is simple, robust, and reliable in our hands, but not all features are yet present.
Tapirs was created by the EvoHull group, at the University of Hull, UK
Detailed instructions for installation, setup, and modification are contained within the Tapirs documentation
- Install conda (miniconda)
- Install git
- Clone the Tapirs repository, and relocate there
git clone https://github.com/EvoHull/Tapirs
cd Tapirs
- Create an environment with snakemake and other software for the workflow
conda env create -f workflow/envs/env.yaml
conda activate tapirs
- Download taxonomy
wget ftp://ftp.ncbi.nih.gov/pub/taxonomy/new_taxdump/new_taxdump.zip
unzip new_taxdump.zip -d resources/databases/new_taxdump
rm new_taxdump.zip
- Populate
resources/databases
with your reference databases andresources/libraries
with your data (a directory containing your demultiplexed R1/R2.fastq.gz sample files) - Place your sample sheet tsv in
config/
(seeconfig/Hull_test.tsv
for layout format) - Adjust
config/config.yaml
to configure the Tapirs workflow (see below) - Dry run
snakemake -npr
to identify any issues - Run
snakemake --cores 4
(you can run all available cores withsnakemake --cores
)
You should adjust config/config.yaml
to specify the location of relevant files (reference databases and sequence data to be analysed) and parameters for the analysis (experiment name, sample sheet name, amplicon/primer lengths, analysis methods etc.). Defaults are present and are set for the test data set: Hull_test.
Consult the Tapirs documentation to get more extensive support.
One example workflow is illustrated below, you may configure yours differently.
EvoHull group, University of Hull, UK
- Dave Lunt (@davelunt)
- Graham Sellers (@Graham-Sellers)
- Michael R Winter (@mrmrwinter)
- Merideth Freiheit (@merfre)
- Marco Benucci