DRAKKAR is a snakemake-based genome-resolved metagenomics pipeline optimised for Mjolnir. Snakemake works along with Slurm to conduct the long pipeline using the optimal memory and time resources. It is built in a modular fashion, so that the entire workflow or only parts of it can be executed. Extended usage tutorial can be found in https://drakkar.readthedocs.io/

Quickstart

module load drakkar/1.0.0 drakkar complete -f input_info.tsv -o drakkar_output

Modules

DRAKKAR is a modular software that allows executing each section of the genome-resolved metagenomic pipeline independently.

• Preprocessing: quality-filters the reads and optionally removes host DNA. drakkar preprocessing {arguments}
• Cataloging: assembles and bins the metagenomic reads using multiple strategies. drakkar cataloging {arguments}
• Annotating: annotates the bins taxonomically and/or functionally, and/or generates community-scale metabolic networks. drakkar annotating {arguments}
• Profiling: conducts genome- or pangenome-based quantitative analyses. drakkar profiling {arguments}

Complete mode

All the modules of DRAKKAR can be run together by using the complete mode. drakkar complete {arguments}

Usage examples

Without sample info file

Minimum usage

drakkar complete -i {input_path} -o {output_path}

-i: path to the folder where the metagenomic sequencing reads are stored. -o: path in which the DRAKKAR outputs will be stored. Metagenomic reads are not mapped to a host genome, individual assemblies are performed, and genome-based profiling is conducted.

With reference genome

drakkar complete -i {input_path} -o {output_path} -r {genome_path}

-i: path to the folder where the metagenomic sequencing reads are stored. -o: path in which the DRAKKAR outputs will be stored. -r: path to the reference genome. Metagenomic reads are mapped to the host genome individual assemblies are performed, and genome-based profiling is conducted.

With reference genome and assembly mode

drakkar complete -i {input_path} -o {output_path} -r {genome_path} -m individual,all -t genomes,pangenomes

-i: path to the folder where the metagenomic sequencing reads are stored. -o: path in which the DRAKKAR outputs will be stored. -r: path to the reference genome. -m: comma-separated list of assembly modes Metagenomic reads are mapped to the host genome, individual assemblies as well as a single coassembly including all samples are performed, and both genome- and pangenome-based profiling is conducted.

With sample info file

|sample|rawreads1|rawreads2|referencename|referencepath|assembly| |---|---|---|---|---|---| |sample1|path/sample11.fq.gz|path/sample12.fq.gz|ref1|path/ref1.fna|assembly1,all| |sample1|path/sample11.fq.gz|path/sample12.fq.gz|ref1|path/ref1.fna|assembly1,all| |sample2|path/sample21.fq.gz|path/sample22.fq.gz|ref1|path/ref1.fna|assembly2,all| |sample3|path/sample31.fq.gz|path/sample32.fq.gz|ref2|path/ref2.fna|assembly2,all| |sample4|path/sample41.fq.gz|path/sample42.fq.gz|ref2|path/ref2.fna|assembly2,all| |sample4|path/sample41.fq.gz|path/sample42.fq.gz|ref2|path/ref2.fna|assembly2,all|

Minimum usage

drakkar complete -f {info_file} -o {output_path} All the required information is extracted from the sample info file.

Minimum usage

drakkar complete -f {info_file} -o {output_path} -m individual Individual assemblies are also conducted on top of the assemblies specified in the sample info file.

DRAKKAR modules

Preprocessing

• Quality-filtering using fastp
• Reference genome indexing
• Reference genome mapping
• Metagenomic and host genomic data outputting

Cataloging

Documentation to be added.

Profiling

Documentation to be added.