aeDNA - a customizable Snakemake workflow for ancient environmental DNA

This workflow combines several modules to build a workflow for ancient environmental DNA (aeDNA) analyses and QC:

• Read trimming
• Read dereplication
• Taxonomy assignment

Authors

• Filipe G. Vieira

Requirements

These workflows require:

• pixi
• conda
• snakemake

Install and configure Snakemake environment

The easiet way to deploy snakemake is to install pixi and use the included environment.

Usage

Step 1: Install and configure workflow

1. Clone this repository to your local system

2. Test pixi environment: pixi run snakemake --version

3. Configure workflow according to your needs via editing the file config/config.yaml.

4. Add samples to analyze:

   • Add samples to config/samples.tsv (column sample is mandatory).
   • For each sample, add one or more sequencing units (runs, lanes, libraries or replicates) to config/units.tsv, as well as adapters used, and path to FASTQ files (if paired-end, use R{Read} to represent R1/R2 files).
   • Missing values can be specified by empty columns or by writing NA.
   • To facilitate things, the helper script make_units.py is provided that, from a list of FASTQ files, generates a folder structure with all necessary files to run (NB: always manually check files before running workflow!).

5. Define your TOKEN

export GITHUB_TOKEN=xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx

Step 2: Test workflow

5. Test your configuration:

5.1. Perform a dry-run:

pixi run snakemake --snakefile <path to Snakfile> --configfile config/config.yaml --dry-run

5.2. Confirm workflow analyses by checking the DAG:

pixi run snakemake --snakefile <path to Snakfile> --configfile config/config.yaml --dag | dot -Tsvg > dag.svg

5.3. Optionaly, create all needed conda environments beforehand:

pixi run snakemake --snakefile <path to Snakfile> --configfile config/config.yaml --jobs 1 --conda-create-envs-only

Step 3: Execute workflow

6. Execute the workflow, using $N cores, either:

6.1. locally:

pixi run snakemake --snakefile <path to Snakfile> --configfile config/config.yaml --jobs $N --software-deployment-method conda

6.2. in a SLURM cluster environment:

pixi run snakemake --snakefile <path to Snakfile> --configfile config/config.yaml --jobs $N --software-deployment-method conda --executor slurm

6.3. Optionally, you can also use binaries from local modules, by adding env-modules to --software-deployment-method:

pixi run snakemake [...] --software-deployment-method env-modules conda [...]

Step 4: Check results

7. After successful execution, you can create a self-contained interactive HTML report:

pixi run snakemake --snakefile <path to Snakfile> --configfile config/config.yaml --report report.html