Skip to content
MetaGEAR

download_databases

One-time setup of the reference databases consumed by every other workflow. Run this once per installation (and re-run when you want to refresh a database).

What it does

Downloads and prepares the following references, sized at tens of gigabytes each:

For microbial_profiles and genes:

  • KneadData host reference (e.g. human genome) for host-read decontamination during QC.
  • MetaPhlAn 4 marker database for taxonomic profiling.
  • HUMAnN 3 ChocoPhlAn (nucleotide) and UniRef90 (protein) databases for functional profiling.
  • GTDB-Tk reference for prokaryotic taxonomic placement of MSPs (gene-analysis only).

For virus:

  • geNomad, CheckV, VirSorter2 databases for viral/plasmid detection.
  • Pharokka database for prokaryotic-virus gene annotation.
  • DRAM-v database for auxiliary metabolic gene (AMG) discovery.
  • iPHoP database for host prediction of viral contigs.
  • AMRFinderPlus database for antimicrobial-resistance gene annotation.

Each tool is downloaded by its native installer wrapped in a Nextflow module; nothing is repackaged.

Inputs

This workflow does not read a samplesheet — there is no --input argument. Instead, you tell it where to put each database via the path parameters below. The metagear-tools CLI manages these paths for you via its config; if you run Nextflow directly you must pass them yourself (typically through a -params-file).

Parameters

ParameterTypeDefaultControls
--databasesstringallWhich database set to download: all, genes, or virus.
--outdirpath(required)Output directory; database files land in <outdir>/<tool>/...
--kneaddata_refdbarray[""]Destination(s) for KneadData host reference(s).
--metaphlan_dbpath—Destination for MetaPhlAn 4 database.
--humann3_nucleopath—Destination for HUMAnN 3 ChocoPhlAn nucleotide database.
--humann3_uniref90path—Destination for HUMAnN 3 UniRef90 protein database.
--gtdb_tk_dbpath—Destination for GTDB-Tk reference.
--genomad_dbpath—Destination for geNomad database.
--checkv_dbpath—Destination for CheckV database.
--virsorter2_dbpath—Destination for VirSorter2 database.
--pharokka_dbpath—Destination for Pharokka database.
--dram_dbpath—Destination for DRAM-v database.
--iphop_dbpath—Destination for iPHoP database.
--amrfinder_dbpath—Destination for AMRFinderPlus database.

Output

Each tool writes into the directory you supplied for its *_db parameter. The pipeline does not republish those files into <outdir> (the modules’ publishDir is intentionally suppressed) — the databases stay where the installers placed them.

Typical layout (paths are whatever you passed):

<metaphlan_db>/                      # MetaPhlAn 4 marker DB
<humann3_nucleo>/                    # ChocoPhlAn pangenomes
<humann3_uniref90>/                  # UniRef90 protein DB
<kneaddata_refdb[0]>/                # Bowtie2 host genome index
<gtdb_tk_db>/                        # GTDB-Tk release
<genomad_db>/                        # geNomad models
<checkv_db>/                         # CheckV references
<pharokka_db>/                       # Pharokka databases
<dram_db>/                           # DRAM-v references
<iphop_db>/                          # iPHoP host genomes
<amrfinder_db>/                      # AMRFinderPlus DB


<outdir>/pipeline_info/              # Nextflow execution metadata

Example

Terminal window
# Download everything (recommended on first install)
nextflow run schirmer-lab/metagear-pipeline -profile docker \
  --workflow download_databases \
  --databases all \
  --outdir databases/ \
  -params-file db-paths.yaml

db-paths.yaml:

metaphlan_db: /data/metagear/metaphlan
humann3_nucleo: /data/metagear/humann/chocophlan
humann3_uniref90: /data/metagear/humann/uniref90_diamond
kneaddata_refdb: [/data/metagear/kneaddata/Homo_sapiens]
gtdb_tk_db: /data/metagear/gtdb_tk
genomad_db: /data/metagear/genomad
checkv_db: /data/metagear/checkv
pharokka_db: /data/metagear/pharokka
dram_db: /data/metagear/dram
iphop_db: /data/metagear/iphop
amrfinder_db: /data/metagear/amrfinder

Notes

  • Disk and time. A full download is in the order of 100+ GB and several hours, mostly bound by external download speed.
  • Run once, point everything at it. Downstream workflows read from these same paths via the same parameters, so once databases are in place all subsequent runs only need --input, --outdir, and -profile.
  • Selective re-runs. Use --databases genes or --databases virus to refresh only one set without re-downloading the other.
  • HUMAnN config. The HUMAnN installer is invoked with --update-config no; HUMAnN’s global config is not modified on the host.
  • Wrapper convenience. metagear-tools ships templates that wire all of these paths into a single user config — recommended over hand-writing the params file above.