HumanCellAtlas / ontology

3 stars 1 forks source link

Help with ontology mappings for mouse endoderm dataset. #39

Closed pnejad closed 5 years ago

pnejad commented 5 years ago

Hi Dani & Zoe,

Could you please help @willrockout and I map fields to the correct ontologies? We took a hack at a few, please let us know if they are incorrect.

MOUSE STRAIN = Afp - GFP

SPECIMEN_FROM_ORGANISM.ORGAN:

SPECIMEN_FROM_ORGANISM.ORGAN_PART

SELECTED CELL TYPE:

daniwelter commented 5 years ago

Afp-GFP - needs to be added to EFO, please submit a new term request ticket. whole embryos - UBERON:0000922 endoderm - UBERON:0000925 gut tube - UBERON:0001555 or UBERON:0007026 (presumptive gut) if it's for an embryo

organ part is not a required field and it doesn't make sense to fill it in for "whole embryo" since by definition the whole "structure" is used, not a sub-part parietal endoderm - UBERON:0008800 visceral and definitive endoderm - UBERON:0004877 (available in full Uberon but not HCAO so needs importing in next release - please file a ticket) and UBERON:0005439 yolk sac endoderm - UBERON:0003257

The selected cell type terms requested don't really make sense to me as they aren't actually specific cell types, just the organ part followed by the word "cell". My understanding of the selected_cell_type field is that it is intended to highlight specific sub-populations of cells that were selected for by some process (eg FACS sorting) in the transition from specimen to cell suspension. I don't think it needs to be explicitly stated that if a specimen comes from the parietal endoderm, then the resulting cell suspension will contain parietal endoderm cells as no specific selection occurred for these cells during the dissociation process. Please provide additional information if my interpretation of the experimental design is incorrect.

zoependlington commented 5 years ago

I will move this ticket to done as the term requests required are in other tickets, but if anything else is needed @pnejad let me know and I'll move this ticket back.