Mohsen Taheri

Results 31 comments of Mohsen Taheri

Use DrugBank test data as a benchmark to evaluate Path Finder.

Use Containment Similarity as the metric, calculated as: ∣ Nodes in one path P ∩ Intermediate nodes ∣ / ∣ nodes in P ∣ #

Use DrugBank test data as a benchmark to evaluate Path Finder.

Drug Bank sample test data: ``` { "PUBCHEM.COMPOUND:118856773": { "KG2_ID": "PUBCHEM.COMPOUND:118856773", "name": "Lepirudin", "category": "biolink:SmallMolecule", "drug_bank_id": "DB00001", "description": "Lepirudin is a recombinant hirudin formed by 65 amino acids that acts...

Use DrugBank test data as a benchmark to evaluate Path Finder.

Hi @dkoslicki, I’m considering using only the core algorithm of PathFinder to test it with this data, without incorporating the Expander results (one hop and two hops, #2398). This approach...

Use DrugBank test data as a benchmark to evaluate Path Finder.

Personal Note: We have 14150 pairs for testing purposes from Drug_bank database.

KG2.10.2c Build

Thanks @bazarkua for letting me know

KG2.10.2c Build

Hi @bazarkua, I have uploaded `curie_ngd_v1.0_KG2.10.2` to `rtxconfig server`

Changelog since 2024-10-17 the deployment to TEST: Sprint 7 changes for Isopod release

Modify the PathFinder service to ensure compatibility with TRAPI requests.

Add Expander one hop and two hops to PathFinder

For more information read this issue: https://github.com/RTXteam/RTX/issues/2369

Add Expander one hop and two hops to PathFinder

Not needed anymore

Make pathfinder progress bar smooth. Get feedback to the user by logging what is going on.

Log paths that excluded due to having general concepts.