Create biosample index to resolve Tissue/cell type entities

[d0choa]

As discussed with @Tobi1kenobi and @DSuveges, we want to create an index containing all the valid tissue/cell type entities for the Open Targets Platform. This will help with validation and resolve labels, ontology, etc.

The most immediate use case is to resolve all the biosample identifiers' eQTL catalogue studies at the API level. In the future, this data can be used for many other use cases: baseline expression, literature-mining, etc. Currently, the QTL study index contains the biosampleId (e.g. UBERON code), but we will rely on the index to retrieve labels or any additional data.

A few tasks:

• [x] create relevant datasource parsers (e.g. UBERON, CL)
• [x] #3478
• [x] create biosampleIndex step to call the parsers and write the single index. Deciding on a reasonable data model.
• [x] develop a qc method in the studyIndex that flags studies with invalid biosample entities
• [x] add the relevant qc flag to the study_validation logic/configuration
• [x] #3544

[Tobi1kenobi]

This has taken a little longer than expected as I initially wrote everything for owl but have since shifted to JSON ontology files. But the code should all be in https://github.com/opentargets/gentropy/compare/alegbe-biosample_index

Would like to test it against the study index currently in gentropy before I make a pull request but unsure of best way to do this

[addramir]

Is it related to tissue/cell line catalogue we previously discussed?

[Tobi1kenobi]

Is it related to tissue/cell line catalogue we previously discussed?

I think so. Long-term it would be great to have unified biosample pages which gather evidence across variants, genes, studies, etc in the same way the variant, gene and study pages do

And an index is certainly necessary for that