Error: Returning missing names

[twoneu]

Hello Dr. Laffy,

I am attempting to use the score function to assign geneset scores to individual cells, but I am encountering the following error:

scrabble::score(mat = as.matrix(seuratobj@assays$RNA@counts), groups = c(geneset1, geneset2, geneset3, geneset4))

Error: Returning missing name(s)...
CCN1
WARS1
MARCHF1
CCN2
CCN1
Error in .check_missing(Group = Group, ref = bins): 
Traceback:

1. scrabble::score(mat = as.matrix(tams_processed@assays$RNA@counts), 
 .     groups = c(tnfa, il2, infg, emt), nbin = 36)
2. binmatch(Group = groups, bins = bins, n = n, replace = replace)
3. .match_bins(Group = Group, bins = bins)
4. .check_missing(Group = Group, ref = bins)
5. stop(cat(c(err, missing), sep = "\n"))

I tried increasing nbins but the error persists. What next steps should I try?

[dyammons]

Hi,
I am not a package developer, but I've been working with this package recently and believe your "returning missing names" error stems from a few of the features in the geneset lists missing from the list of features in your Seurat count matrix.

The "clean data" steps in the code below should help you avoid the error. It is easiest to do these steps if your gene lists are in a dataframe with long format.

Best,
Dylan

#load Seurat data
rawData <- as.matrix(seuratobj@assays$RNA@counts)

#get example getset lists into df
genesetList <- list("geneset1" = c("CD3E", "CD8A", "CD4"),
                    "geneset2" = c("MS4A1", "PAX5", 'CD79A'),
                    "geneset3" = c("ANPEP", "CD68", "MHCII"),
                    "geneset4" = c("CD34", "SELL", "ICAM")
)

genesetList_df <- gather(as.data.frame(do.call(cbind, genesetList)))
colnames(genesetList_df) <- c("geneSet", "gene")

#clean data
genesetList_df <- dplyr::filter(genesetList_df, gene %in% rownames(rawData))
genesetList_df <- genesetList_df[!duplicated(genesetList_df$gene), ]

#get back into list
geneSetList <- genesetList_df %>% split(x = .$gene, f = .$geneSet)

######## CALC ENRICHMENT SCORE ##############

#get score
enrichment_score <- scrabble::score(mat=rawData,
                          groups=geneSetList,
                          binmat = NULL,
                          bins = NULL,
                          controls = NULL,
                          bin.control = FALSE,
                          center = TRUE,
                          nbin = 30,
                          n = 100,
                          replace = TRUE
                         )