Code maintenance, documentation and improvements

• unit testing
  • Test for the formatting of command lines from each node is correct
  • Check that the consequence of a given node on the data written on the disk are appropriate (e.g. LR-QC is expected to exclude some stacks and this exclusion is supposed to be taken into account in the downstream nodes)
• check for the proper installation of dependencies
• Reproducibility tests
  • Use the simulated data
  • Run the full pipeline on this dataset across all institutions in order to ensure that we have the exact same results. This might be important for instance to ensure that different versions of docker/singularity images do not influence the results.
  • Need to implement small quantitative measures to compare across institutions, would be direct one we have implemented the feature extraction nodes.