In the ontology generation framework, scripts named OWLNETS-UMLS-GRAPH (aka "Jonathan's scripts") add data from the OWLNETS output of the PheKnowLator script to the base set of UMLS CSVs that are imported into an ontology neo4j.
The original script was developed as a Jupyter Notebook. The original development workflow was:
Jonathan modified the notebook.
A HuBMAP analyst ran a "transform.py" script to convert the notebook to a "pure python" script.
One consequence of this workflow is the presence of files with names like OWLNETS-UMLS-GRAPH-n.py.
Part of the reason for this workflow is the difficulty of using Jupyter with github.
I (Alan) stopped using this workflow. The final version of the file is the one named OWLNETS-UMLS-GRAPH-12.py. This version of the script will be the source of truth that is used, especially in github. I work directly within the pure python script; Jupyter is only used for debugging.
Issue
In the ontology generation framework, scripts named OWLNETS-UMLS-GRAPH (aka "Jonathan's scripts") add data from the OWLNETS output of the PheKnowLator script to the base set of UMLS CSVs that are imported into an ontology neo4j.
The original script was developed as a Jupyter Notebook. The original development workflow was:
One consequence of this workflow is the presence of files with names like OWLNETS-UMLS-GRAPH-n.py. Part of the reason for this workflow is the difficulty of using Jupyter with github.
I (Alan) stopped using this workflow. The final version of the file is the one named OWLNETS-UMLS-GRAPH-12.py. This version of the script will be the source of truth that is used, especially in github. I work directly within the pure python script; Jupyter is only used for debugging.
Solution
To prevent later confusion,