metamash

Snakemake and Snaketool pipeline to taxonomically profile ONT long read metagenomics.

Snakemake and Snaketool pipeline to taxonomically profile ONT long read metagenomics with sourmash

Much code and ideas have been taken and adapted from the pb-metagenomics-tools repository.

Additionally this pipeline has been substantially informed by this benchmarking study.

Assumes that your reads are in a directory, fastq.gz suffixed.

Installation

git clone "https://github.com/gbouras13/metamash.git"
cd metamash/
pip install -e .
metamash --help
metamash install --help
metamash run --help

Steps:

Download the CHM13 human-t2t-hla.fa host genome (from hostile) with the following command.

metamash install --database host_genome_db

Run trimnami specifying the directory of FASTQ reads as --reads ensuring the mapping is for ONT

trimnami run --reads metamash/test_data --host host_genome_db/human-t2t-hla.fa --minimap map-ont --output trimnami_output nanopore

You will find the host depleted and trimmed reads in <output>/nanopore so here e.g. trimnami_output/nanopore.

Download the mmseqs2 v 13.45111 uniref50 database - this takes ages in Australia~

e.g.

# will write to 'uniref50' directory
mmseqs databases UniRef50 uniref50 tmp --threads 3