Website would be better if it showed info about tissue origins of cell types

[NathanSkene]

E.g. while discussing with Prof Meyr, he wanted to know what tissue the "AFP ALB positive cells" come from... couldn't give him anything on this. Maybe we should have a way to click through to this info?

Maybe this will be less of a problem with the cell type names from the new dataset

[bschilder]

That information is available here:

• https://github.com/neurogenomics/rare_disease_celltyping/blob/master/data/DescartesHuman_celltype_mapping.csv

Many cell types occur in many tissues. I can certainly add this info, but may need to figure out a way to reasonably display that info. Could be a hover item or a separate table below the results.

[bschilder]

Some of the celltypes defined in DescartesHuman were unique to that dataset, thus marker-based names like "AFP ALB positive cells". It's all described in the supp info of their paper, but i can add the Cell Ontology aligned definitions as well to provide better interpretability.

[bschilder]

Instead of manually assigned cell types to the tissue we think they should belong to (introduces bias), I used the metadata from the original scRNA-seq datasets and included it as a data object in MSTExplorer:

MSTExplorer::tissue_maps

You can then use this to compute the proportion of each cell type in each tissue, or the tissue in which each cell type is most commonly found:

map <- MSTExplorer::map_tissue()