semsim
Details
| GitHub | Knowledge-Graph-Hub/semsim |
| Language | Jupyter Notebook |
| Description | Write out file with semantic similarity of all pairs of ontology terms in a given ontology |
Dependencies
External Dependencies
| Package | Version |
|---|---|
| python | ^3.9 |
| grape | ^0.1.10 |
| pandas | ^1.4.3 |
| tox | ^3.25.1 |
| click | ^8.1.3 |
| tqdm | ^4.64.0 |
Documentation
Write out file with semantic similarity of all pairs of HP terms
Workflow: - download HPOA file in order to calculate frequencies for each term - download HPO - turn HPO into DAG - drop singletons - drop non-HP terms - confirm DAG-ness - load graph into Ensmallen - make Resnik model, calculate all by all Resnik similarity - make Jaccard model, calculate all by all Jaccard similarity - pandas join tables - write out to file - upload file to S3 (KG-Hub?) # decide on this later