Implement proper SSSOM validation

[matentzn]

Fixes #30 and #31

[glass-ships]

should we also add metadata for mondo?
it seems to already be in the header, but:

    curie_map:
      BFO: http://purl.obolibrary.org/obo/BFO_
      CSP: http://purl.obolibrary.org/obo/mondo/mappings/unknown_prefix/CSP/
      DECIPHER: http://purl.obolibrary.org/obo/mondo/mappings/unknown_prefix/DECIPHER/
      DERMO: http://purl.obolibrary.org/obo/mondo/mappings/unknown_prefix/DERMO/
      DOID: http://purl.obolibrary.org/obo/DOID_
      EFO: http://www.ebi.ac.uk/efo/EFO_
      GARD: http://purl.obolibrary.org/obo/mondo/mappings/unknown_prefix/GARD/
      GTR: http://purl.obolibrary.org/obo/mondo/mappings/unknown_prefix/GTR/
      HGNC: http://identifiers.org/hgnc/
      HP: http://purl.obolibrary.org/obo/HP_
      ICD10CM: http://purl.bioontology.org/ontology/ICD10CM/
      ICD10EXP: http://purl.obolibrary.org/obo/mondo/mappings/unknown_prefix/ICD10EXP/
      ICD10WHO: https://icd.who.int/browse10/2019/en#/
      ICD9: http://purl.obolibrary.org/obo/mondo/mappings/unknown_prefix/ICD9/
      ICD9CM: http://purl.obolibrary.org/obo/mondo/mappings/unknown_prefix/ICD9CM/
      ICDO: http://purl.obolibrary.org/obo/mondo/mappings/unknown_prefix/ICDO/
      IDO: http://purl.obolibrary.org/obo/IDO_
      IEDB: http://purl.obolibrary.org/obo/mondo/mappings/unknown_prefix/IEDB/
      LOINC: http://purl.obolibrary.org/obo/mondo/mappings/unknown_prefix/LOINC/
      MEDGEN: http://identifiers.org/medgen/
      MESH: http://identifiers.org/mesh/
      MFOMD: http://purl.obolibrary.org/obo/MFOMD_
      MONDO: http://purl.obolibrary.org/obo/MONDO_
      MP: http://purl.obolibrary.org/obo/MP_
      MPATH: http://purl.obolibrary.org/obo/MPATH_
      MTH: http://purl.obolibrary.org/obo/mondo/mappings/unknown_prefix/MTH/
      MedDRA: http://identifiers.org/meddra/
      NCIT: http://purl.obolibrary.org/obo/NCIT_
      NDFRT: http://purl.obolibrary.org/obo/mondo/mappings/unknown_prefix/NDFRT/
      NIFSTD: http://purl.obolibrary.org/obo/mondo/mappings/unknown_prefix/NIFSTD/
      OBI: http://purl.obolibrary.org/obo/OBI_
      OGMS: http://purl.obolibrary.org/obo/OGMS_
      OMIM: https://omim.org/entry/
      OMIMPS: https://omim.org/phenotypicSeries/PS
      OMOP: http://purl.obolibrary.org/obo/mondo/mappings/unknown_prefix/OMOP/
      ONCOTREE: http://purl.obolibrary.org/obo/mondo/mappings/unknown_prefix/ONCOTREE/
      Orphanet: http://www.orpha.net/ORDO/Orphanet_
      PATO: http://purl.obolibrary.org/obo/PATO_
      PMID: http://purl.obolibrary.org/obo/mondo/mappings/unknown_prefix/PMID/
      RO: http://purl.obolibrary.org/obo/RO_
      Reactome: http://purl.obolibrary.org/obo/mondo/mappings/unknown_prefix/Reactome/
      SCDO: http://purl.obolibrary.org/obo/SCDO_
      SCTID: http://identifiers.org/snomedct/
      UMLS: http://linkedlifedata.com/resource/umls/id/
      Wikidata: http://purl.obolibrary.org/obo/mondo/mappings/unknown_prefix/Wikidata/
      Wikipedia: http://purl.obolibrary.org/obo/mondo/mappings/unknown_prefix/Wikipedia/
      oboInOwl: http://www.geneontology.org/formats/oboInOwl#
      orphanet.ordo: http://www.orpha.net/ORDO/Orphanet_
      owl: http://www.w3.org/2002/07/owl#
      rdf: http://www.w3.org/1999/02/22-rdf-syntax-ns#
      rdfs: http://www.w3.org/2000/01/rdf-schema#
      semapv: https://w3id.org/semapv/vocab/
      skos: http://www.w3.org/2004/02/skos/core#
      sssom: https://w3id.org/sssom/

[matentzn]

You mean because of the "unknowns"? In this case, make a MONDO issue about it - they should fix it!

[glass-ships]

ah. no, i didn't even notice the "unknowns". I just meant that #30 seemed to want curie_map added for all mapping sets listed in the registry, but this PR only did so for biomappings and gene mappings

[matentzn]

Mondo already has it, right, so no need to add another one?

[matentzn]

You have merged this but qc was failing! (Because sssom py cannot process gene_mappings). In any case @hrshdhgd is trying profile the validation run to figure out
If this is fixable.

[glass-ships]

reverted

[matentzn]

It's fine to merge the PR, in any case nothing was more broken afterwards then it was before.. it was just a general warning not to merge stuff if qc fails

[matentzn]

Are you making a new PR?

[glass-ships]

as of this commit, f589a59 the tests were passing, at least as far as i can tell?

[glass-ships]

Are you making a new PR?

Trying to, but git doesn't seem to think there's any changes between this branch and main anymore, so I'm not sure.

[matentzn]

Huh whaaaa?? Sorry man. Seems all good then! I didn't get a message it's all fixed!

[matentzn]

You may have to revert the revert

[hrshdhgd]

The QC passes but it still took 50 mins to pass. sssom was not added as a dependency in the toml file and that's why it was failing in the first place. I don't know why it was thinking for over an hour.

[matentzn]

No it definetly had sssom installed, as it finished the pipeline parts before that - but maybe an older version?

[hrshdhgd]

I set up locally from scratch and sssom was not in the toml file. I added it and that led to qc passing.

[hrshdhgd]

Don't revert anything. I will make a new PR if need be.

[matentzn]

Please don't lose all my changes 😅 thanks

[hrshdhgd]

haha don't worry. Nothing will be lost.