Tibanna runs portable pipelines (in CWL/WDL/Snakemake/shell) on the AWS Cloud.
Install Tibanna.
pip install tibanna
Use CLI to set up the cloud component and run workflow.
# Deploy Unicorn to the Cloud (Unicorn = serverless scheduler/resource allocator).
tibanna deploy_unicorn --usergroup=mygroup
# Run CWL/WDL workflow on the Cloud.
tibanna run_workflow --input-json=myrun.json
Alternatively, use Python API.
from tibanna.core import API
# Deploy Unicorn to the Cloud.
API().deploy_unicorn(usergroup='mygroup')
# Run CWL/WDL workflow on the Cloud.
API().run_workflow(input_json='myrun.json')
Note: Starting 0.8.2
, Tibanna supports local CWL/WDL files as well as shell commands and Snakemake workflows.
Note 2: As of Tibanna version 2.0.0
, Python 3.6 is no longer supported. Please switch to Python 3.8! Python 3.7 is also supported as a fallback, but please prefer 3.8 if you can.
Note 3: Starting 0.8.0
, one no longer needs to git clone
the Tibanna repo.
- Please switch from
invoke <command>
totibanna <command>
! - We also renovated the Python API as an inheritable class to allow development around tibanna.
For more details, see Tibanna Documentation.
- Also check out our paper in Bioinformatics.
- A preprint can also be found on biorxiv.