NCATSTranslator / TACT

For voting on two-month release cycle items
0 stars 3 forks source link

Pause QA of chembl-based results for MVP1 until CTKP swap out is complete #14

Open mbrush opened 4 months ago

mbrush commented 4 months ago

Copying this comment from Feedback repo issue 805, as I think it is a topic for TACT to discuss.

I think we should pause on testing/reporting any issues for MVP1 related to results based on chembl and their ingest of clinical trials data. It has become clear from talking with KPs and ARAs that teams are at different stages or implementing the treats refactor, and some inconsistencies remain in how it is being done. On top of this, we know that these refactored chembl ingests will all be replaced by the CTKP in the next month or so anyway. Trying to QA during this transition period is IMO not a good use of our time - given that the CTKP swap-out will result in things looking very different (in a good way - cleaner, less duplication, less inconsistency) - and many issues that testers are painstakingly documented will be fixed as a result of this swap-out.

I propose that QA be directed to MVP2 over the next month, and we make a concerted push get the CTKP swap out complete ASAP. Once this has been done and made its way into test, we can shift testing focus back to MVP1.

I am happy to help coordinate the moving parts here - KPs needing to swap out their chembl ingests, Gwenlyn/Multiomics in their CTKP implementation and move over to automat, the CQS team in ensuring that 'treats' predictions continue to get generated correctly, and proper EPC reported after the swap.

Maybe we can discuss on the next TACT call - and agree on a plan that will be most efficient here?