Closed gaurav closed 3 years ago
I think this is going to be about using the CEDAR Workbench or some other tool to validate metadata being ingested into CDS, as well as answering the big question about whether this validation should be centralized (at CCDH or somewhere else in CRDC) or whether each Node would be responsible for its own ingest and validation, using standards and possibly tools developed by CCDH. In either case, we should probably figure out the minimum metadata first (#16), and then figure out how to build a validator around that. So I'm postponing this to Q4 this year.
We should figure out which quarter we want to do this in. I want to say "next quarter" -- we should have all the transformation and testing stuff working in our GitHub repositories this quarter, so we'll be ready to think about turning these into services and UIs for others to use then. But I have also lost track of what quarter we're currently in :).
At the moment, it looks like we'll probably carry out validation with our own LinkML-based tools (probably using JSON Schema or ShEx) rather than using the "Metadata Validator", which I think is a reference to the GDC MVS (https://github.com/CBIIT/GDCMVS). I'm going to close it for now. If we need this functionality in the future, we can open a new issue that has a clear use-case as to precisely who needs this functionality and why. (Please re-open this issue if I'm wrong!)
Scheduled: Q3 2020, Q4 2020