MGI gene to pub ingest - Githubissues

monarch-initiative / monarch-ingest

Data ingest application for Monarch Initiative knowledge graph using Koza

https://monarchinitiative.org

15 stars 2 forks source link

MGI gene to pub ingest #168

Closed glass-ships closed 2 years ago

glass-ships commented 2 years ago

Create ingest of MGI Gene to Publication data.

Note: this is still on version 0.1.2 of Biolink Model Pydantic.

When that's updated to latest, we'll want to update all tests/scripts to:

Change NamedThingToInformationContentEntityAssociation to InformationContentEntityToNamedThingAssociation
Switch the order of objects and subjects in "gene_to_publication" ingest files
- Does this make them "publication_to_gene" ingests now?
Change predicate.mentions to predicate.mentioned_by

kevinschaper commented 2 years ago

Here's the extras that I think it still needs:

Add an entry to download.yaml for the file
Some tests (I don't think anything inventive, just following the same pattern of confirming that entities come out for the row?)
Add the ingest to the ingest list in monarch_ingest/ingest_pipeline.py
Add a markdown doc for the ingest in the docs/ directory

glass-ships commented 2 years ago

Curious about something I noticed looking through the output of this ingest - it looks like for rows of genes with multiple publications, the way it's written, there will be duplicate gene nodes. As many as there are publications that mention that gene.

Is that the normal structure for a nodes file? Or should the writer ignore duplicate node?