nf-core/metatdenovo taxonomy

The data in this repository can be used to assign taxonomy to sequences with Diamond [Buchfink et al. 2015], particularly using the --diamond_dbs parameter in nf-core/metatdenovo, release 1.1 or later.

Currently, the data available represents species-representative genomes from the Genome Taxonomy Database (GTDB), release R09-RS220 [Parks et al. 2018].

File preparation

All species-representative genomes from GTDB were downloaded from the National Center for Biotechnology Information (NCBI) and annotated with Prokka [v. 1.14.6; Seemann 2014], and the sequences for all resulting proteins were used for this data. The taxonomy dump files (in NCBI taxonomy dump format) were created from the GTDB metadata with TaxonKit [v. 0.18.0; Shen and Ren 2021] and the Diamond database with Diamond [v. 2.1.10; Buchfink et al. 2015] in "taxonomy mode", i.e. using the taxonomy dump created with TaxonKit. (See below for commands used.)

File descriptions

There are five files:

1. gtdb-r220.faa.gz: Fasta file with protein sequences. Not used by nf-core/metatdenovo but can be used to create the Diamond database below.
2. gtdb-r220.taxonomy.dmnd: Diamond database with taxonomy information.
3. gtdb-r220.names.dmp: Taxonomy dump file.
4. gtdb-r220.nodes.dmp: Nodes dump file.
5. gtdb-r220.seqid2taxid.tsv.gz: Mapping from protein accession to taxon.

The Diamond database and taxonomy dump files can be used with nf-core/metatdenovo (Version >1.1) by providing a csv file like below to the --diamond_dbs parameter. (Although Nextflow can use https-urls for paths, it is usually better to download the very large files and keep local copies.)

db,dmnd_path,taxdump_names,taxdump_nodes,ranks,parse_with_taxdump

gtdb,gtdb_r220_repr.dmnd,gtdb_taxdump/names.dmp,gtdb_taxdump/nodes.dmp,domain;phylum;class;order;genus;species;strain,

Commands used to prepare taxonomy dump files and the Diamond database

• Taxonomy dump: cut -f 1,19-20 *metadata.tsv | grep -v 'accession' | awk 'BEGIN { FS="\t" } { if ( $2 == "t" ) { print $1 "\t" $3 } }' | taxonkit create-taxdump --gtdb -O .
• Diamond database: gunzip -c gtdb-r220.faa.gz | sed '/^>/s/ .*//' | diamond makedb --taxonmap gtdb-r220.seqid2taxid.tsv.gz --taxonnames gtdb-r220.names.dmp --taxonnodes gtdb-r220.nodes.dmp --db gtdb-r220.taxonomy.dmnd --no-parse-seqids

Revision history

20250211 First version