You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
If you run RenameCells on a seurat object with one cell (I'll put below why this came up in my analysis), it does not actually rename the cells because of a <= not < in RenameCells.Assay (see here). It's possible that this <= is required but git blame at least doesn't help me understand why it was added.
The reason why this came up is I had a list of Seurat objects, one cell type per object. I ultimately wanted a single seurat object of cell type averages so first AverageExpression'd each element of the list (to get a Seurat object with one 'cell') and then was going to merge together. It was impractical to first merge all the seurat objects, then average by Ident, because they were so large they hit R's 2^31-1 size limitations. However, AverageExpression leaves each cell's ID as just all so all the cell ID's were the same pre-merge. When merge.Seurat tries to make the cell IDs unique, it uses RenameCell which then silently fails and later processing errors. I can't manually unique-ify the cell names for similar RenameCell reasons.
I ended up just extracting the average matrices myself and rbind'ing but it's a pretty pernicious error :-)
Thanks
Jonah
The text was updated successfully, but these errors were encountered:
Hi,
If you run
RenameCells
on a seurat object with one cell (I'll put below why this came up in my analysis), it does not actually rename the cells because of a<=
not<
inRenameCells.Assay
(see here). It's possible that this<=
is required butgit blame
at least doesn't help me understand why it was added.The reason why this came up is I had a list of Seurat objects, one cell type per object. I ultimately wanted a single seurat object of cell type averages so first
AverageExpression
'd each element of the list (to get a Seurat object with one 'cell') and then was going to merge together. It was impractical to first merge all the seurat objects, then average by Ident, because they were so large they hit R's 2^31-1 size limitations. However,AverageExpression
leaves each cell's ID as justall
so all the cell ID's were the same pre-merge. Whenmerge.Seurat
tries to make the cell IDs unique, it usesRenameCell
which then silently fails and later processing errors. I can't manually unique-ify the cell names for similarRenameCell
reasons.I ended up just extracting the average matrices myself and
rbind
'ing but it's a pretty pernicious error :-)Thanks
Jonah
The text was updated successfully, but these errors were encountered: