Wrong grounding of gastric cancer publications to gonorrhoea

[d0choa]

A user in community reported:

We observed that many of the papers tagged with gonorrhea are actually about gastric cancer and never mention the word gonorrhea. On a related note, or perhaps as a result, many of the genes associated with gonorrhea are oncogenes associated with gastric cancer. Given that gonorrhea and gastric cancer are not related but both start with the letter “g”, we are wondering if there is a systematic error in paper-disease or gene-disease relationship tagging.

https://community.opentargets.org/t/literature-disease-association-error-papers-tagged-with-gonorrhea-are-really-about-gastric-cancer/1453

First diagnosis from @DSuveges points to our grounding pipeline

[d0choa]

I believe the problem is that MONDO correctly has GC as a synonym for gonorrhoea.

However, people in the manuscripts occassionally refer to gastric cancer as GC, but we don’t have GC as a synonym for Gastric Cancer.

When we find a GC tagged as a disease, our grounding will unequivocally try to ground it as ghonorrea. If we were to have the GC as a synonym/acronym of gastric cancer, our disambiguation would kick in and it would most likely get the right answer.

@zoependlington would it be possible to consider GC as a synonym of Gastric Cancer?

[zoependlington]

@d0choa We could request adding GC as a synonym in Mondo as the term is a Mondo term, but it would be up for their consideration whether to add or not.

[d0choa]

Feel free to cross-reference the ticket. If necessary, we can build a strong case based on all the occurrences of GC as an abbreviation of gastric cancer in the biomedical literature (similar to the screenshot above).

[zoependlington]

Requested here: https://github.com/monarch-initiative/mondo/issues/7828