Consistency data endpoints

[matthijsbrouwer]

I did some checks on the consistency of the output from the BrAPI endpoints:

callsets

• for 58 entries the callSetDbId is 'None' (the string with 4 letters) : this is probably not the correct identifier, and of course it is not unique
• the returned studyDbId do not correspond to the studyDbId used in the studies endpoint : for example the studyDbId 'Agent project' is returned, where the studies endpoint uses integer studyDbId values and does not contain 'Agent project'

samples

• not all the returned germplasmDbId correspond to a germplasmDbId used in the germplasm endpoint : for example 'GB20_SSD_TRI_10886_3901' is returned, and this is not in the germplasm endpoint
• for 42460 entries the observationUnitDbId is 'None' (the string with 4 letters)
• the returned studyDbId do not correspond to the studyDbId used in the studies endpoint : for example the studyDbId 'Agent project' is returned, where the studies endpoint uses integer studyDbId values and does not contain 'Agent project'

attributevalues

• the returned attributeDbIds are all string values, but the attributeDbId in the attributes endpoint are all integers

germplasm

• the used commonCropNames are 'Barley', 'Lettuce', 'Wheat', but only two lowercase(!) versions can be found from the divbrowse endpoints: 'wheat' and 'barley'

variables

• commonCropName is not provided; if this does make sense, it is probably a good idea to provide it

studies

• commonCropName is not provided; if this does make sense, it is probably a good idea to provide it

---

variants

• there are referenceDbIds returned (only for barley), but there is no references endpoint
• there are referenceSetDbId returned (only for barley), but there is no referencesets endpoint

Implementing references and referencesets does seem to make sense, because here also commonCropName can be placed to distinguish between variants/callsets for barley and wheat