MODIFICATIONS

Kraken is no longer able to download automatically the files from the NCBI database because they rearranged the structure of all the RefSeq genomes. Two possible solutions are described here:
a) Change the paths in the script download_genomic_library.sh – for instance, for bacteria (line 43) we use  $FTP_SERVER/genomes/archive/old_refseq/Bacteria/all.fna.tar.gz [ref]
b) Create our own databases. To make it easier, we can use the scripts here [ref]. The magic in these scripts is that they also account for the suppression of the GI number that the NCBI has also adopted [ref].
c) Use the script UpdateDatabasesKraken.py created by Sej Modha, which is now in the /tools/kraken folder.
In addition, we had to install jellyfish v1.1 (Kraken does not support v2) from the website where it is originally stored http://www.cbcb.umd.edu/software/jellyfish/