Include submission or creation date in ClinVar evidence

[apriltuesday]

OT is interested in when the connection between a variant(hence gene) and a phenotype/disease become public.

• Check whether we can use CreatedDate of the RCV or whether we need to parse the SCV and use the oldest submission
• Agree on update to JSON schema
• Implementation (in a separate issue if SCV parsing is needed, otherwise should be easy enough to do in this one)

[apriltuesday]

@tcezard @M-casado Observations on the first point:

RCV creation date is usually a few months after the oldest submission date, but does sometimes stray further and sometimes precedes the first submission (perhaps from a submission being withdrawn?):

The dates are within a year of each other 75% of the time.

All this aside, I think RCV creation date is the thing OT is looking for, especially on consulting ClinVar's schema definition: DateCreated is the date when the record first became public in ClinVar.

[eric-czech]

FWIW, we have some related discussion on this in https://community.opentargets.org/t/enable-evidence-trend-analysis-e-g-in-clinvar/1235/4

[apriltuesday]

Hi @eric-czech, thanks for this and especially the document from ClinVar, it's really useful! It's good to hear of others thinking along similar lines, and also seeing the kinds of questions people using the data are interested in :+1: