Finer GC trait mapping may justify aggregation methodology

[jeremyjyang]

EFO revisions and finer trait mappings have greatly increased the number of traits. This has value but may justify methods for aggregating by combining closely related traits. Some details from GC support (Elizabeth Lewis, 28-Sep-2023):

Our EFO colleagues have been working over the last year to make several major improvements that impact on the GWAS Catalog, including:

1. Improving the structure of EFO, particularly in the disease and measurement branches
2. Improving the workflow to scale up the number of new term requests that can be processed

For the first improvement, the team are working on aligning measurement traits with OBA (Ontology of Biological Attributes). Work is ongoing and expected to be completed before the end of the year, at which point the entire measurement branch will be replaced with new terms and structure. This is a big upgrade due to the flatness of that branch currently. We were working closely with the EFO team on this and will publicise and create documentation when the changes occur.

For the second improvement, this has been necessary due to large protein, metabolite and biobank papers which can report thousands of GWAS at a time. We are now able to create specific terms for each metabolite (for example), whereas previously we might have used a parent term, which explains the huge growth in number of traits.

You may wish to contact efo-users@ebi.ac.uk if you would like more information ahead of time.