OxBreaker

Nextflow implementation of the pipeline described in OxBreaker: species-agnostic pipeline for the analysis of outbreaks using nanopore sequencing. The pipeline takes a set of reads, sequenced using Oxford Nanopore Technologies (ONT), and determine whether they are closely related.

The aim of this pipeline is to enable healthcare professionoals and Infection Prevention and Control (IPC) teams idenfity incipient outbreaks in their hospitals, and manage the available resources in case of no suspected outbreak.

The implementation found here works in computers running GNU/Linux or macOS with Nextflow, and its associated dependencies installed, and has the following syntax and options:

-entry oxbreaker

Nextflow program to be run.

-profile hpc|standard

Sets whether the pipeline uses Slurm workload manager, common in research high performance computing (HPC) environments, or not. standard is the preferred option if OxBreaker is run on a local workstation.

--input STR

Full path containing the location of the .fastq.gz files to analyse.

--output STR

Full path where the results from the pipeline will be saved.

--reference STR

Full path of the file containing the reference genome, in genbank format. This parameter is REQUIRED if the Krake2 Microbial Database (30GB) is not used.

--db STR

Full path of the directory containing the Krake2 Microbial Database (30GB). This parameter is REQUIRED if no reference genome is provided.

--min_mq INT

Minimum mapping quality below which OxBreaker will discard the reads, in PHRED format ranging 0-60 (defaults to 55).

--min_read_number INT

Minimum supporting reads for OxBreaker to consider a variant with respect to the reference genome (defaults to 10).

--min_freq FLOAT

Minimum variant frequency for OxBreaker to consider a variant with respect to the reference genome, ranging from 0.0 to 1.0 (defaults to 0.9).

-phylogeny

Produce a phylogeny as part of the outputs of OxBreaker.

--debug

Root phylogeny on the reference given, useful to check if the results are meaninful.

Experimental

OxBreaker uses bcftools as backend for variant-calling. But it can also use Clair3. However, given the current state of Clair3 for bacterial genomes, these settings remain experimental as further studies are run:

--clair3

Enables the use of Clair3 to find variants with respect the reference genome provided.

--model STR

Full path of the directory containing the model used to train Clair3 based on chemistry and other ONT characteristics. This parameter is REQUIRED if -clair3 is used.

--chemistry R9|R10

Specify chemistry used for the ONT sequencing run (defaults to R10).

Sample script

The code below is an example of how to run locallyl OxBreaker, with default settings, adding a phylogeny rooted in this_reference.gb as part of the output files:

#!/usr/bin/bash
READS_DIR=/directory/containing/data
OUT_DIR=/directory/project/output

    # Run OxBreaker
    nextflow run main.nf -resume \
    -entry outbreaker \
    -profile standard \
    --input $READS_DIR \
    --output $OUT_DIR \
    --reference /directory/references/this_reference.gb \
    --phylogeny --debug

Carlos Reding © 2026. University of Oxford.