Skip to content

Mapper for Mondo and MeSH using Leonmap#7

Open
HarshitSoni1903 wants to merge 3 commits intogyorilab:mainfrom
HarshitSoni1903:main
Open

Mapper for Mondo and MeSH using Leonmap#7
HarshitSoni1903 wants to merge 3 commits intogyorilab:mainfrom
HarshitSoni1903:main

Conversation

@HarshitSoni1903
Copy link
Copy Markdown

No description provided.

@buzgalbraith
Copy link
Copy Markdown
Contributor

buzgalbraith commented Apr 7, 2026

Hey @HarshitSoni1903 great work on this!

It would be really helpful if we could filter down the mappings in scripts/leonmap_mondo_mesh_predictions.sssom.tsv so we could have an apples to apples comparisons with other mapping systems we have looked at.

So you do not need to re-run the system, you can just add post-processing logic to scripts/generate_leonmap_mondo_mesh_mapping.py such that it:

  1. Drops any values where the MONDO terms starts with obsoloete. This prefix implies the term is deprecated and no longer actively in use. An example of the type of mapping we may want to drop for this is obsolete deafness on line 14 of the mapping file.
  2. Try to match the novelty filter used in scripts/generate_mondo_mesh_mappings.py. At a high level what this is doing is filtering the predicted mappings to only those with "one to one novelty", this means that neither of the terms are used in any other mappings.

Feel free to let me know if you have any questions!

@HarshitSoni1903
Copy link
Copy Markdown
Author

Updated the code with suggested changes. The script now does:

  1. A complete two-way pass to ascertain truly novel mappings.
  2. Offloading task to Mapnet's utils module, 3 files generated:
    • Right: Ontologies that are rightly mapped
    • Wrong: Ontologies that are either mapped to some other ontologies in known mappings
    • Novel: Truly new mappings, Source and Target ontologies that do not have a known mapping

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

2 participants