aponsero / HPC_KrakenFor16S

HPC pipeline for 16S taxonomic profiling using Kraken2

HPC_KrakenFor16S

Bash pipeline for 16S taxonomic profiling using Kraken2. Developed for NBI cluster (SLURM scheduler).

Pipeline overview:

The pipeline takes raw or QC'd amplicon reads and run Kraken2 and optionally Bracken for rapid taxonomic profiling. An optional merging step will parse the Kraken2/Bracken output and generate a PhyloSeq object. Grey steps in the workflow below are optional steps in the pipeline.

flowchart LR
    A[Raw reads] -.|Cutadapt|.-> B[trimmed reads]
    B --> C{Kraken2}
    C -->D[Kraken2 profiles]
    C -.|Bracken|.-> E[Bracken profiles]
    D -.|Merge|.-> F[PhyloSeq]
    E -.|Merge|.-> G[PhyloSeq]

Note that the pipeline takes paired end files that follow the naming convention:

- PAIR1 = ${FILE}_R1_001.fastq.gz
- PAIR2 = ${FILE}_R2_001.fastq.gz

Requirements

This pipeline requires:

- Kraken2 v2.1.3
- Bracken v2.9
- Cutadapt v4.7
- R v4.2.1 (and the Phyloseq, tidyverse and r-argparse packages)

The required tools can be installed using conda and the provided .yml environment description.

conda env create -f env/KrakenFor16S.yml

Run the pipeline

Edit the config.sh file

Select the steps to run and specify the input and output folder.

activate the conda environment and run script

conda activate KrakenFor16S
bash run_KrakenFor16S.sh