brianraymor
commented
1 year ago There may be an additional case where fields could be hidden in Explorer. This is related to visium fields that are added to .obs in squidpy modeling. Currently, the curators are stripping the fields prior to submission - BUT, if we wanted to retain the fields for dataset downloaders to reuse in squidpy or seurat when we add official support, we may want to hide the fields in Explorer.
The following tissue_positions.csv fields are removed .obs ['in_tissue'], .obs ['array_row'], .obs ['array_col'].
CC: @niknak33
seve
There may be fields in the underlying data that would be helpful to data downloaders but overwhelm Explorer's visualization capabilities. We'd like to hear form the community if this functionality would be helpful.
So far we have heard from Kyle at the Allen Institute: "The files we send to you have significantly reduced metadata so as to not overwhelm the left hand panel in your viewer. This has caused some confusion for users who download the data from your portal, instead of from our AWS Open Data Registry (which has the full metadata available)."