update BioThings semmeddb: download files ASAP?

[colleenXu]

Semmeddb had a final update sometime after May 8, 2024 (underlying pubmed data goes up to this date). VS BioThings semmeddb's build date is 2024-04-04, so it isn't using the latest data.

Andrew and I agreed that it'd be good to update this pending API (Repo?)

The website seems to imply that it might be archived this month, so maybe the needed files need to be downloaded ASAP?

CC @everaldorodrigo @ctrl-schaff

This probably also involves updating the NGD endpoint data and generating the metatriple document count file

---

Potentially useful previous issues:

• https://github.com/biothings/semmeddb/issues
• NGD endpoint PR? NGD implementation for semmeddb2 #121
• update new semmeddb api #63
• update SemMedDB APIs #30

[NikkiBytes]

The current parser works with the updated data, I did not run into any parser issues, so I think the new data integration should be straight forward. @newgene
Here is an example doc:

{
  "_id": "C0028613-INTERACTS_WITH-83985",
  "predicate": "INTERACTS_WITH",
  "predication": [
    {
      "predication_id": 87884415,
      "pmid": 28202305,
      "sentence_id": 48848743,
      "sentence": "We previously reported that LbNH, a nucleoside hydrolase from Lactobacillus buchneri LBK78 (NITE P-01581), was the first enzyme found to act on 2'-OMe-NRs.",
      "subject_text": "nucleoside hydrolase",
      "subject_score": 983,
      "object_text": "NRs",
      "object_score": 1000
    }
  ],
  "pmid_count": 1,
  "predication_count": 1,
  "subject": {
    "umls": "C0028613",
    "name": "Nucleosidases",
    "semantic_type_abbreviation": "aapp",
    "semantic_type_name": "Amino Acid, Peptide, or Protein",
    "novelty": 1
  },
  "object": {
    "ncbigene": "83985",
    "name": "SPNS1",
    "semantic_type_abbreviation": "gngm",
    "semantic_type_name": "Gene or Genome",
    "novelty": 1
  }
}

[colleenXu]

What about updating the NGD endpoint data and generating the metatriple document count file?

[NikkiBytes]

@colleenXu Yes! Confirming no updates will be need for the parser itself!

[NikkiBytes]

@newgene @colleenXu the metatriple documents were generated (based off the notebook guide) . I passed over the files to you @colleenXu last week, but from the workflow here I think it is completed, the output looks good, lets confirm that. I updated the notebook a little, for efficiency, but if we confirm those files look good this is wrapped up. From my understanding, the NGD endpoint will be inspected once we push this forward. The parser code updates themselves are very simple, a few filenames will be updated, nothing more. Also we need to decide where we store the metatriple document file.
Next steps....

• Inspect and confirm metatriple file -- decide where to store the file
• Push minor code parser updates to repo (what exact repo should I push to?)
• Once updated we can inspect the NGD endpoint

[NikkiBytes]

@colleenXu @newgene the new files have been added and the API is updated. [link here] If everything checks out, we can close this issue.