Guppy is a data processing toolkit that contains the Oxford Nanopore Technologies' basecalling algorithms, and several bioinformatic post-processing features.

See this tutorial: https://denbi-nanopore-training-course.readthedocs.io/en/latest/basecalling/basecalling.html

• Version: 6.2.1
• Added: July, 2021
• Updated: July, 2022

Guppy is only available on compute06 because this is the only node that has a GPU.

Note: guppy ships with some pre-configured models that set many basecalling parameters to sensible defaults. Please consult: /opt/ont/guppy/data.

Notes from the sysadmin during installation.

$ sudo dnf install https://cdn.oxfordnanoportal.com/software/analysis/ont-guppy-6.2.1-1.el8.x86_64.rpm

Run some test data using fourteen CPUs:

$ mkdir -p /var/scratch/aorth/2022-05-24-guppy
$ cd /var/scratch/aorth/2022-05-24-guppy
$ wget https://openstack.cebitec.uni-bielefeld.de:8080/swift/v1/nanopore_course_data/Data_Group1.tar.gz
$ tar xf Data_Group1.tar.gz
$ guppy_basecaller --compress_fastq -i data/fast5_tiny/ -s data/basecall_tiny/ --cpu_threads_per_caller 14 --num_callers 1 -c dna_r9.4.1_450bps_hac.cfg
...
Caller time: 3281435 ms, Samples called: 27619692, samples/s: 8416.96

Run some test data using a GPU (currently only on compute06):

$ srun --gres=gpu:v100:1 --partition=batch -w compute06 -n 1 -J guppy --pty bash
$ guppy_basecaller --device "cuda:0" --compress_fastq -i data/fast5_tiny/ -s data/basecall_tiny/ -c dna_r9.4.1_450bps_hac.cfg
...
Caller time: 4366 ms, Samples called: 27619692, samples/s: 6.32609e+06