Closed kevinschaper closed 7 months ago
This is merged and deployed to beta.monarchinitiative.org, with deployment to production to follow in the next day or two
I'm wondering: how complex is your implementation?
For example: do all the edges with the same subject-category/subject-namespace/predicate/object-category/object-namespace have the knowledge level / agent_type info?
I think right now they're just fixed for each of our ingests, so even less granular than that.
Thanks for the quick reply!
I'm wondering if I can set a knowledge_level / agent_type for each of those combos for BTE/Service Provider's use in its annotation of this API, in the short-term. It sounds like I can but...
I'm not sure if each combo only has one ingest.
Here's a distinct + count on subject_category, subject_namespace, predicate, object_category, object_namespace, knowledge_level, agent_type, provided_by (provided_by being where I'm sticking the ingest name) that might be useful. It's a tsv, but GitHub doesn't like the tsv extension, so renamed it to csv.
definitely still fixed per ingest though:
knowledge_level | agent_type | provided_by | count_star() |
---|---|---|---|
logical_entailment | automated_agent | hpoa_gene_to_phenotype_edges | 308455 |
knowledge_assertion | manual_agent | zfin_gene_to_phenotype_edges | 149558 |
knowledge_assertion | manual_agent | ctd_chemical_to_disease_edges | 5644 |
knowledge_assertion | not_provided | panther_genome_orthologs_edges | 551212 |
knowledge_assertion | not_provided | reactome_chemical_to_pathway_edges | 68572 |
knowledge_assertion | manual_agent | go_annotation_edges | 2591280 |
knowledge_assertion | manual_agent | alliance_gene_to_phenotype_edges | 304589 |
not_provided | not_provided | phenio_edges | 775320 |
knowledge_assertion | manual_agent | alliance_gene_to_expression_edges | 1879092 |
knowledge_assertion | not_provided | bgee_gene_to_expression_edges | 436178 |
knowledge_assertion | not_provided | reactome_gene_to_pathway_edges | 203937 |
knowledge_assertion | manual_agent | hpoa_gene_to_disease_edges | 15327 |
knowledge_assertion | manual_agent | dictybase_gene_to_phenotype_edges | 1216 |
knowledge_assertion | manual_agent | pombase_gene_to_phenotype_edges | 169665 |
knowledge_assertion | not_provided | biogrid_edges | 1418164 |
knowledge_assertion | manual_agent | hpoa_disease_mode_of_inheritance_edges | 8535 |
knowledge_assertion | manual_agent | hpoa_disease_to_phenotype_edges | 245981 |
knowledge_assertion | not_provided | string_protein_links_edges | 1475505 |
knowledge_assertion | manual_agent | xenbase_gene_to_phenotype_edges | 2232 |
Noticed for this line: the source says hpoa, but the actual edges' sources say omim/medgen
biolink:Gene HGNC biolink:causes biolink:Disease MONDO knowledge_assertion manual_agent hpoa_gene_to_disease_edges 6707
To meet a Translator milestone we need to populate Knowledge Level and Agent Type on all of our edges in the 2024-04 Release.
Here are some guidelines: https://github.com/NCATSTranslator/ReasonerAPI/blob/master/ImplementationGuidance/Specifications/knowledge_level_agent_type_specification.md
With the short turnaround, the goal will be to populate easy to find values first and at least be able to validate by setting others to
not provided
- also to prioritize edges in our Gene/Disease/Phenotype triangle.