Try to run workflow B.1 and not have it explode

[finnagin]

query: https://github.com/NCATSTranslator/minihackathons/blob/main/2021-12_demo/workflowB/B.1_DILI-three-hop.json

1. For edge e1 filter on attribute provided by COHD and ICEES
2. Go through edge by edge experimenting with overlaying fisher exact, Jaccard, and ngd as well as filtering on those
3. Experiment with adding in is_set = True

[finnagin]

@dkoslicki does this issue do a good job of capturing what you were requesting at the OW meeting?

[dkoslicki]

Yes, that does capture it. Note that the goal is to tamp down the huge number of edges/results without sacrificing quality/relevancy. So point 2 is more like: "experiment" with overlays...

[edeutsch]

Workflow B.1 seems to run fine with the recent changes and tests, so I think this can be closed? @finnagin ?