Minutes_Standards_2023 03

Standards Call 2023-03

Review bugs and other issues for v1.4.2 patch release (#674).
Discuss Germline related issues in #663, #680, #667, and #652. Including clarifying how best to have the various Germline schema be integrated into the data model without being required, whether IARC/OGRDB needs an x-airr attribute for custom nullable/required flags, and how to handle de-nesting.
Do we need to clarify the contribution criteria and licensing?

#663: Suggested fix is to change the template code to default enums to empty string for type string. Fall back to the first value if that doesn’t work. We also need to understand use case for validating template file. We can use the miairr attribute as is to clarify nullable status w.r.t. IARC with miarr.default = defined, essential = required/non-nullable, and important = required/nullable.
#667: By default, objects within the schema do not contain objects on their own (this is a consensus on style, not a hard requirement by the schema). This is currently not met by some Germline object (and a couple of other ones, e.g. ontology.top_node). Should be fixed.
Leave documented_alleles, undocumented_alleles and deleted_genes separate. Rename undocumented_alleles.allele_name to label (likely in v2.0).
We can denest Subject.genotype (move receptor/mhc up). Questions are outstanding as to whether this should be an array and whether the genotype should live in Repertoire instead of Subset. We will have to wait until v2.0 to address these as they will break compatibility for iReceptor.