Eric Deutsch

Thanks @amykglen ! I deployed it to devED but am having some problems. New code deployed: ``` ISSUE_TEMPLATES/kg2rollout.md | 8 +++++--- code/ARAX/ARAXQuery/ARAX_expander.py | 29 ++++++++++++++++++++++++----- code/ARAX/ARAXQuery/Expand/trapi_querier.py | 12 +++++++++++- code/ARAX/test/test_ARAX_expand.py...

maybe a logging message includes a QEdge object instead of a string?

ah yes, here, I think: https://github.com/RTXteam/RTX/blob/15fabe9fe40c5e4dd911515fc398831bc3f26653/code/ARAX/ARAXQuery/Expand/trapi_querier.py#L93 I will try fixing that unless you wave me off..

yes! I commented out that line and that fixed it. Overall I think this patch is a good bandaid. Before: https://arax.ncats.io/test/?r=252443 (2 results) After: https://arax.ncats.io/devED/?r=252445 (6 results) I think this...

it probably didn't cause an error during pytest because the logging data isn't serialized to JSON there. It was during the JSON serialization that things went bad. perhaps a more...

this is great, thanks! I think think your proposed annotation is good and fine. others can take issue with it if they notice it and bring it up. I think...

just rolled out to arax2.ncats.io

In KG2.10 there are no more _treats_ only _applied_to_treat.... The "patch" maps these predicates to _treat_ for non-Semmed edges Question is: With Drug Approvals KP now avail, do we back...

Hi @chunyuma thank you for your analysis. Unfortunately a lot of things things changed and this is probably responsible for the why the performance did not increase like I expected....

Note that there are different versions of code running on arax.ncats.io. https://arax.ncats.io/test is running on the `master` branch, but https://arax.ncats.io/ is running from the `production` branch, which is substantially different...