This example is a tumour-normal processing workflow applied to a
publically available whole-genome sequencing (WGS) dataset. As this
example uses WGS data, a large amount of data will be downloaded.
Furthermore, a large amount of space will be required to build the final
products.

# Building on local machine

Run `nix build -I bionix=../..` in this directory.

# Building via HPC (slurm or torque)

Run `nix build -f cluster.nix` to build on slurm. Note that Nix must be
configured such that the temporary build directories are created on
shared storage.

For Torque, run `nix build -f cluster.nix -I bionix=../.. --argstr
tmpDir /scratch/`. Unlike the slurm handler, a shared tmpdir location
must be specified.

In both cases, you may need to adjust the resource limits specified in
cluster.nix to suit your particular cluster hardware.