kraken2-paired-reads.md

Classify Paired-End Reads Using Kraken2

The kraken2.nf nextflow script will take any number of paired reads in FASTQ format and classify reads using Kraken2.

Dependencies (version tested)

• Nextflow (24.04.4)
• Java (18.0.2.1)
• Python (3.10)
• Kraken2 (2.1.3)

Conda Environment

Create environment using conda:
conda env create -f ./nextflow-pipelines/env/kraken2.yml

Activate conda environment:
conda activate kraken2 or source activate kraken2

Usage

#!/bin/bash

# Activate conda environment
conda activate kraken2

# Variables
cpus=20
reads=/path/to/reads/directory
krakenDB=/path/to/krakenDB/
outdir=/path/to/output/directory

# Run pipeline
nextflow run ~/nextflow-pipelines/src/kraken2.nf \
    --reads ${reads} \
    --suffix "_{1,2}.fq.gz" \
    --krakenDB ${krakenDB} \
    --outdir ${outdir} \
    --cpus ${cpus}

Parameters	Description
--reads	path to input directory containing FASTQ files
--suffix	string denoting the suffix after a sample name and the forward (read1) and reverse (read2) designation (e.g. for read pair sampleID_1.fq.gz and sampleID_2.fq.gz set the parameter to --suffix "_{1,2}.fq.gz". The name of this output file will be called sample.kraken)
--krakenDB	path to Kraken2 database
--kraken2	string of additional arguments passed to kraken2 (e.g. "--gzip-compressed --minimum-hit-groups 3")
--outdir	path to output directory
--cpus	integer denoting the number of cpus (default: 16)

Input

$ ls input/
SampleID_01_1.fq.gz SampleID_02_1.fq.gz
SampleID_01_2.fq.gz SampleID_02_2.fq.gz

Output

$ ls output/
SampleID_01.kraken SampleID_02.kraken
SampleID_01.report SampleID_02.report