To install kraken2 on SCG (or any other machine), you should create a conda environment with all the requirements. See the scg_tools repo for more information.
Clone this repository, then use the commands below to install. You must first cd to the location where you cloned the repository.
It is also tasteful to copy datasets.tsv and config.yaml to a working directory for modification and workflow execution.
# crete an environment classification2 with all the requirements
conda env create -f classification2.yaml
# update taxonomy used with Krona plotting software
# not necessary for just classification
mkdir $(which kraken2 | sed 's/envs\/classification2.*$//g')/envs/classification2/bin/taxonomy
ktUpdateTaxonomy.sh