/ontology-mapper-evaluation

Resources to reproduce the comparison of ontology mappings generated by text2term and selected benchmarks

Primary LanguagePython

ontology-mapper-evaluation

Resources to reproduce the comparison of ontology mappings generated by text2term with:

Running the comparison code

The mapping comparison can be executed by running the Python script:

python compare_ontology_mappings.py

Inputs

The inputs for the comparison are:

Benchmark mappings

  • data/UK_Biobank_master_file.tsv: UK Biobank phenotype mappings, downloaded on 2024-03-14 (URL).
  • data/biomappings.tsv: Biomappings table, downloaded on 2024-04-08 (URL).
  • data/efo.ols.sssom.tsv: mappings of various ontologies to EFO, downloaded on 2024-04-10 (URL).

Target ontology

  • data/efo_edges.tsv: table containing all pairs of asserted subsumptions between terms in EFO.
  • data/efo_entailed_edges.tsv: table containing all pairs of entailed subsumptions between terms in EFO.

Outputs

The outputs of the comparison are, for each benchmark mapping set:

  • output/{dataset}_t2t_mappings.csv: table containing the generated text2term mappings.
  • output/{dataset}_mappings.tsv: table containing mappings extracted from the original dataset.
  • output/{dataset}_results.tsv: table containing the results of the mappings' comparison.