Seurat: Trouble subsetting data according to barcode suffix

[xihuiyin]

Hi, so I have aggregated samples in one barcode file which the end of each cell is denoted by "-1" for sample 1, "-2" for sample 2, and so on. I'm trying to create separate objects out of the suffixed barcodes, but I can't find a way to do it. I've also tried adding metadata by adding a separate column by the sample name and then reading it again, but I keep getting the error: "Error in .rowNamesDF<-(x, value = value) : invalid 'row.names' length" at the 4th line of code.

rownames = colnames(gd)
write.csv(rownames,"gddata.csv")
condition.only.metadata <- read.csv(file = "gddata.csv", header = FALSE)
rownames(condition.only.metadata) = rownames([email protected])
gd <- AddMetaData(object = gd, metadata = condition.only.metadata, colnames = "condition")

I've also tried

gd.matrix <- gerd@assays[["RNA"]]@counts 
colnames(gd.matrix) <- paste0(colnames(gd), "-6") 
nodctrl3 <- CreateSeuratObject(gd.matrix)

But I quickly realized that this code adds additional suffixes at the end of each cell, instead of actually filtering based on the suffix and creating a separate object.

Is there a better way to do this?

[samuel-marsh]

Hi,

When you read into Seurat as single object does the orig.ident column in your metadata not match the barcode suffixes? It should normally. Then you can just subset using:
object_sample01 <- subset(object_all, subset = orig.ident == "1")

Alternatively, you could filter the input csv file first before creating Seurat object.

Best,
Sam

[xihuiyin]

I figured out the problem (one of my columns was unlabelled) but I ran into more issues. However I managed to figure them out. Thank you so much for your help!

[vkavaka]

You can extract sample number from rownames and put it into the separate column of meta.data file like this:
[email protected]$samples <- sub('.*-', '', rownames(gd))
Then subset aacording to sample number:
Idents(gd) <- 'samples'
Sample1 <- subset(gd, idents = '1')