search

Knowledge-Graph-Hub / kg-phenio

A Graph for experiments doing ML on ontologies.
BSD 3-Clause "New" or "Revised" License
6 stars 4 forks source link

Add config file to specify transform parameters and pass deprecated flag #121

Closed caufieldjh closed 7 months ago

caufieldjh commented 8 months ago

See example config: https://github.com/biolink/kgx/blob/master/examples/sample-transform-config.yml

caufieldjh commented 8 months ago

This currently works - if kgx finds a deprecated: true, then it adds it to the subsets column of the tsv. The koza transform then parses that as a node attribute with the id owl:deprecated.

Here's the problem: in the transformed PHENIO tsv, only 84 nodes out of 287404 get this attribute. The deprecated flag appears in phenio.json 17086 times and 86 times in the first kgx tsv transform, so kgx is clearly missing or ignoring most of the flags. This is because it doesn't appear to pass through the deprecated flag at all, but some are getting through because they're also labeled as being part of deprecated subsets.

caufieldjh commented 8 months ago

This may be blocked until https://github.com/biolink/kgx/issues/470 is complete since the deprecated column just isn't making it into the tsv output after the json -> kgx tsv transform.