Michael Baudis

@jrambla Well, the definition's `default: false` - if implemented correctly - prevents faulty matches. The description says: > queried property has an excluded property However, it is difficult to avoid...

@david4096 So: Implementing `svtype` etc. as 1st class named fields in `Variant` (like the current `end`)? Or `VariantInfo` object as collector for all named variant things beyond basics? I am...

> how do we handle the fact that alternate bases can be multiply valued if we are only providing the single sv annotation @david4096 Not sure what you mean by...

@david4096 So for structural variants (in contrast to small indels), no "alternate_bases" are provided; just the type of alteration (e.g. DUP, DEL); position (= `start`) and end (INFO:END in VCF;...

@david4096 No; here we follow VCF, which would put the type of variant into the `ALT` column (e.g. as ``). Not saying this is optimal (yes, I get the "confusing");...

@david4096 This variation of the example * changes `alternate_bases` to an empty list * adds `variant_type` * moves `cipos`, `ciend` out of `info` * keeps `svlen` in `info`, since it...

@jjfarrell (@david4096, @KyleGao, @kozbo ...) Thanks for this comment/background/opinion! Fortunately, this is in line with my take ... Now this can be decomposed into different points: 1. getting rid of...

This has been taken up in PR https://github.com/ga4gh/schemas/pull/772 .

@M-casado Just modify & PR against schema-urgent-fixes. AFAIK (not sure where this came from) this has no constraining intention and just misses an ellipsis in the examples (but better to...

> Thank you for the PR, I think it has great additions. Nonetheless, I always feel that `RefSeqId`, `Assembly` and now `VariantId` are very open, that the only restriction is...