varianttype and alternateBases and matename should use oneof construct in OpenAPI specs

[blankdots]

Related to https://github.com/ga4gh-beacon/specification/issues/273 and https://github.com/ga4gh-beacon/specification/issues/277#issuecomment-483760720

Seems like:

• alternateBases or variantType cannot be used with mateName;
• if a query specifies variantType=BND it seems like it is a looks mateName should not be present.

Or am i missing something?

[mbaudis]

You need mateName=chro... + variantType=BND to specify a fusion event with known partners.

Sent from my iPad

On Apr 24, 2019, at 9:00 AM, Stefan Negru [email protected] wrote:

Related to #273 and #277 (comment)

Seems like:

alternateBases or variantType cannot be used with mateName; if a query specifies variantType=BND it seems like it is a looks mateName should not be present. Or am i missing something?

— You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHub, or mute the thread.

[blankdots]

@mbaudis the API description leaves room for interpretation thus my confusion, maybe worth detailing this particular thing in a shape or form. But just to be clear if one would specify mateName shouldn't variantType be assumed as BND by default, or are there any cases we are missing?

[mbaudis]

@blankdots You’re correct; but IMO a well-formed fusion query should specify BND, even if it is implicit. I’m not sure how to go further about this - specification or documentation for query types - on a technical level; a “oneOf” construct should in principle apply to query prototypes, not single attributes. @sdelatorrep?

[blankdots]

ok, open to any suggestion how to specify, oneof was a proposal. Also for an implementation standpoint it would be another check that needs to be performed that when mateName is used with variantType to be sure that any service querying the API used the correct value: BND - and that is ok, just highlighting that this might be useful when clarifying in the API specs.