Guppy
Guppy is a data processing toolkit that contains the Oxford Nanopore Technologies’ basecalling algorithms, and several bioinformatic post-processing features. It is provided as binaries to run on Windows, OS X and Linux platforms, as well as being integrated with MinKNOW, the Oxford Nanopore device control software.
Guppy provides guppy_aligner, guppy_basecaller and guppy_barcoder, and can run on GPUs and CPUs.
Due to the software license, we cannot install Guppy in our central software location, but it is simple to download, expand and then use in your jobs.
Example
Guppy is distributed as a tar archive, and is available to ONT customers via their community website https://community.nanoporetech.com/.
The file normally has a filename like ont-guppy_version_linux64.tar.gz (GPU version) and ont-guppy-cpu_version_linux64.tar.gz (CPU version).
Download the relevant file and transfer it to your project directory on Spartan.
The Guppy tar archive can be expanded using tar. In the following example, we have transferred the guppy archive to /data/gpfs/projects/punim0001/guppy
[scrosby@spartan-login1 ~]$ cd /data/gpfs/projects/punim0001/guppy
[scrosby@spartan-login1 guppy]$ ls
ont-guppy_6.1.2_linux64.tar.gz
[scrosby@spartan-login1 guppy]$ tar -zxf ont-guppy_6.1.2_linux64.tar.gz
[scrosby@spartan-login1 guppy]$ ls
ont-guppy ont-guppy_6.1.2_linux64.tar.gz
This has expanded ont-guppy_6.1.2_linux64.tar.gz into the directory ont-guppy
To use Guppy in a job, just reference the commands using their full path. e.g. to use guppy_basecaller that you expanded into /data/gpfs/projects/punim0001/guppy/ont-guppy, add the following to your Slurm submit script
/data/gpfs/projects/punim0001/guppy/ont-guppy/bin/guppy_barcoder: error while loading shared libraries: libcuda.so.1: cannot open shared object file: No such file or directory
You receive this error if you run the GPU version of Guppy on a CPU node. Either run it on the GPU partitions, or use the CPU version of Guppy.