ancestries' ontology and /rest/ancestry_categories

[pgscatalog]

ancestries' ontology and /rest/ancestry_categories

[ens-lgil]

Dear @ramiromagno,

When we implemented the "Display category", at the beginning this was only meant to be used in the pie charts, to provide a summarised information about the ancestry distribution for each Score.
We also updated the "ancestry_broad" data so now they all match the NHGRI-EBI GWAS ancestry categories.

Then when we did some users testing, it appeared that there was an interest to have these distributions available in the metadata exports and, by extend, in the REST API. However we didn't intend to provide a detailed mapping of the distribution with the "Ancestry categories".

Regarding the rest/score endpoints, we choose to use a 2/3 letters notation to reduce the size of the rest call (mainly for the rest/score/all endpoint) and keep the endpoints as fast as we could (we store the distribution in the database in a similar structure, therefore this is relatively quick to build and return the REST API call).

The /rest/ancestry_categories endpoint is mainly there to provide the labels corresponding to the 2/3 letters ancestry notations.
However, thanks to your suggestions, we are thinking about improving the /rest/ancestry_categories endpoint by providing the "Ancestry categories" associated with the "Display categories", e.g.:

"AFR": {
	"display_category": "African", 
	"categories": [
		"African American or Afro-Caribbean", 
		"African unspecified",
		"Sub-Saharan African"
	]
},
"EUR": {
	"display_category": "European",
	"categories": [
		"European"
	]
},
...

Best regards,
Laurent

[ramiromagno]

Thank you @ens-lgil for the update.