Kraken_db_install_scripts

Updated Kraken DB install scripts to cope with new-ish NCBI structure

Dependencies

• Kraken installed and in your $PATH
• bioperl

Usage

Usage is very similar to adding a custom database as explained in the Kraken manual.

First, download the taxonomy:

DB_NAME="kraken_db"
kraken_shell_scripts/kraken-build --download-taxonomy --db $DB_NAME

This step will download taxonomy tree info and the GI to taxid mapping file. Since we're manually adding a new library, the GI's will not be used, and the accessions numbers will be mapped to the taxids at the next steps

Then, download the genomes you want to add the your kraken database. With the perl scripts from this directory, you can download:

• complete genomes from refseq:

  • archaeal genomes
  • bacterial genomes
  • fungal genomes
  • protozoan genomes
  • viral genomes

• Homo_sapiens.GRCh38 (primary assembly)

If you want to install the complete viral genomes from RefSeq, per example:

perl download_viral.pl
find viral/ -name '*.fna' -print0 | \
    xargs -0 -I{} -n1 kraken_shell_scripts/kraken-build \
    --add-to-library {} --db $DB_NAME
kraken_shell_scripts/kraken-build --build --db $DB_NAME

License

The scripts under kraken_shell_scripts are GPL3
The rest is MIT