RenameCell error when replacing cell-names

[swvanderlaan]

For purposes of anonymization I need to rename the cell-names which include the study number in our dataset.

I have a Seurat-object with test@assays$RNA$counts@Dimnames[2]) and test@assays$SCT$counts@Dimnames[2]) having for instance: "aab1.P2.299" "aab1.P2.311" "aab1.P2.315" "aab1.P2.318" "aab1.P2.319" "aab1.P2.321" "aab1.P2.322" "aab1.P2.338" "aab1.P2.349".

Here's what I did.

First, I got the list of original cell names in a new column:

test@meta.data$orig.cellname <- unlist(c(test@assays$RNA$counts@Dimnames[2]))

Next, I split this column:

test@meta.data <- test@meta.data %>% separate(orig.cellname, c('Cell_SNR', 'Cell_Plate', 'Cell_nr'))

Then I created a new cellname:

test@meta.data$new.cellname <- paste(test$SampleID, test$Cell_Plate, test$Cell_nr, sep = ".")

Where SampleID is a new anonymized sample ID.

Finally, I apply RenameCell.

test <- RenameCells(test, new.names = test@meta.data$new.cellname)

Here I assume the dimnames will be replaced throughout the SeuratObject, so at test@assays$RNA$counts, test@assays$RNA$data, test@assays$SCT$counts, and test@assays$SCT$data, but also at test@graphs$SCT_nn, and test@graphs$SCT_snn.

I get this error:

Error in new.cell.names.global[match(x = working.cells, table = old.names)] <- new.cell.names : 
  NAs are not allowed in subscripted assignments

I don't get this as far as I can see there aren't any 'NAs' in my new column.

I am not sure how to move forward from here. Any advice or suggestions would be great.