search

clingen-data-model / clingen-interpretation

Allele (variant) interpretation model and API for ClinGen
3 stars 1 forks source link

Do we need _RelatedReferenceSequence and _RelatedContextualAllele? #131

Closed larrybabb closed 6 years ago

larrybabb commented 6 years ago

@bpow @cbizon I'm not sure how these extended data points are critical to our interpretation model. I do think that they may be interesting in regards to the Allele model, but they seem to just generate noise and produce maintenance here.

Can we drop them?

cbizon commented 6 years ago

Where do they come from?

larrybabb commented 6 years ago

At some point early on we captured the relationsihp between a transcript and a protein sequence, similar to how we captured some of the related contextual variants. In the end, I don't see any of this information making up into our Statement (Info) types in the context. It's possible we thought we would need to re-model portions of the allele model but then moved in a different direction.

I also don't see it in the produced web documentation.

I say we toss it for now. We can always re-institute it if needed.

cbizon commented 6 years ago

Sorry I am still confused. What class are you talking about?

bpow commented 6 years ago

These have not shown up in the examples because I wasn't linking them to their relevant referrants (perhaps they got shuffled into different sheets at some time and I didn't notice...)

There is now code to handle these sheets in the middleman project ( commit 839b581 )

bpow commented 6 years ago

Just to add... removing these would only affect the examples, but not the VCI translator code (since it calls out to the allele registry for allelic representations.

larrybabb commented 6 years ago

Replaced by adding 2 attributes to ContextualAllele and ReferenceSequence called "produces" and "producedBy" which map to the only 2 relation types that we have defined to date.