rename error when use RenameGenesSeurat

[songlyzz]

Description
Hi, vertesy:
When I use the function of RenameGenesSeurat, I got some errors that :
RNA@data is of type dgeCMatrix!
Error in validObject(object = x) :
The object with the category "LogMap" is incorrect: Rownames cannot be empty strings

Can you help me？I post my code below.

THANKS，
SONG

for (sample in folders ){ 
  sc_object <- CreateSeuratObject(
    counts = Read10X(file.path(sample), gene.column = 1),
    project = sample,
    min.cells = 3, min.features = 200
  )
  
  # rename 
  ids=mapIds(EnsDb.Mmusculus.v79,
             keys = rownames(sc_object), # the keys to select records for from the database.
             keytype="GENEID",
             columns="SYMBOL",
             column="SYMBOL")
  ids=as.data.frame(ids)
  ids$ENSEMBL=row.names(ids)
  ids=ids[!duplicated(ids$ids),]
  ids=ids[!duplicated(ids$ENSEMBL),]
  pos=match(ids$ENSEMBL,rownames(sc_object) )
  sc_object=sc_object[pos,]
  
  obj =RenameGenesSeurat(obj = sc_object,
                         newnames = ids$ids)
  sc_object<-obj[,-grep("NA",rownames(obj))]
}

[vertesy]

Hi Song, sure, happy to,

could you please tell me your version of:

1. Seurat
2. Seurat.utils
3. the seurat version of sc_object.

This bug seems to be zombie: i solved it mutiple times, but keeps coming back

[songlyzz]

Hi ,
I changed my code like this:

ids=select(org.Mm.eg.db,keys = rownames(sc_object),
             columns = c('ENSEMBL','SYMBOL'),
             keytype = 'ENSEMBL')
  head(ids)
  dim(ids) # 
  ids=na.omit(ids)
  dim(ids) 
  length(unique(ids$SYMBOL)) # 
  ids=ids[!duplicated(ids$SYMBOL),]
  ids=ids[!duplicated(ids$ENSEMBL),]
  pos=match(ids$ENSEMBL,rownames(sc_object) )
  sc_object=sc_object[pos,]
  
  sc_object =RenameGenesSeurat(obj = sc_object,
                         newnames = ids$SYMBOL)

This time it do well !

THANKS,
SONG

[vertesy]

great!

[songlyzz]

Hi vertesy,
Now I have an another problem that in the org.Mm.eg.db, I did not have like "mt-" mitochondrial genes, how can I filter it when the name is not standard for PercentageFeatureSet(sc_object, pattern = "^mt-") function that in order to QC.

THANKS，
SONG

[vertesy]

I am sorry but this goes beyond what I can support.
PercentageFeatureSet is anyways Seurat, not this package

[songlyzz]

Hi vertesy:
Thank you very much! I know why my first method is not work, because here is a empty symbol gene name to match the ensembl id, but the org.Mm.eg.db database matches it, so the second one work well.
I can fix them together to solve the problem.

THANKS，
SONG