Asset 392: bad answer ID for DNMT inhibitor

[colleenXu]

Asset 392 says the expected output name is DNMT inhibitor. However, the output ID CHEMBL:4034633 is not a proper CURIE and appears to be an assay/document ID, not a chemical entity.

Using the Chembl API, I find this info on the document but it doesn't mention what the chemical used is (click here)

The Target ID is also odd and seems to lack info.

{
  "aidx": "941973",
  "assay_category": null,
  "assay_cell_type": null,
  "assay_chembl_id": "CHEMBL4034633",
  "assay_classifications": [],
  "assay_organism": "Homo sapiens",
  "assay_parameters": [],
  "assay_strain": null,
  "assay_subcellular_fraction": null,
  "assay_tax_id": 9606,
  "assay_test_type": null,
  "assay_tissue": null,
  "assay_type": "B",
  "assay_type_description": "Binding",
  "bao_format": "BAO_0000219",
  "bao_label": "cell-based format",
  "cell_chembl_id": null,
  "confidence_description": "Default value - Target unknown or has yet to be assigned",
  "confidence_score": 0,
  "description": "Inhibition of DNMT in human U937 cells assessed as induction of P16 CpG islands demethylation at 1 to 2.5 uM after 36 hrs by PCR analysis",
  "document_chembl_id": "CHEMBL4033641",
  "relationship_description": "Default value - Target has yet to be curated",
  "relationship_type": "U",
  "src_assay_id": null,
  "src_id": 1,
  "target_chembl_id": "CHEMBL612545",
  "tissue_chembl_id": null,
  "variant_sequence": null
}

If this test is kept, I imagine the expected output ID needs to be changed.

[colleenXu]

Related to other issues:

• Too general assets #64?
• I think all tools have been consistently failing this test: Tests that no one passed #69

[maximusunc]

Name Resolver suggests this curie: UMLS:C2746050. Thoughts?

[gaurav]

That does seem like the best identifier, although it's a single-identifier clique and a UMLS term. I would have loved a CHEMBL or CHEBI term, but I can't find one.

[colleenXu]

Perhaps the original writer of the asset could help? Maybe a specific DNMT inhibitor chemical could be used as the answer?

[maximusunc]

This asset doesn't have a writer specified, but an educated guess is @sstemann ?