MetacellsByGroups -> "Error in apply(cell_sample, 1, function(x) { : dim(X) must have a positive length"

[MJD-BlanckGroup]

Hi all,

Currently running into some difficulty generating metacell groupings. I am working with a relatively small single-cell dataset (dim = 23684 x 3515), but did not expect to run into this much trouble... unable to generate metacells, despite having the pca reduction input, and confirmed "seurat_clusters" metadata loaded into the seurat_obj@meta.data slot. Might anyone have any insight into this error code?

Warning: Data is of class dgeMatrix. Coercing to dgCMatrix.
Error in apply(cell_sample, 1, function(x) { :
dim(X) must have a positive length

Thanks much in advance!!

dat <- read.table("dat, log2tpm.csv", sep = ",", header = 1, row.names = 1)

dat <- (2^dat - 1)
seurat_obj <- CreateSeuratObject(counts = dat, project = "SingleCellProject", assay = "RNA")
if (!inherits(seurat_obj@assays$RNA@layers[["counts"]], "dgCMatrix")) {
  seurat_obj@assays$RNA@layers[["counts"]] <- as(seurat_obj@assays$RNA@layers[["counts"]], "dgCMatrix")
}
seurat_obj <- NormalizeData(seurat_obj)

seurat_obj <- FindVariableFeatures(seurat_obj, selection.method = "vst", nfeatures = 2000)
seurat_obj <- ScaleData(seurat_obj)
seurat_obj <- RunPCA(seurat_obj, features = VariableFeatures(object = seurat_obj))
seurat_obj <- FindNeighbors(seurat_obj, dims = 1:10)
seurat_obj <- FindClusters(seurat_obj, resolution = 0.5)
seurat_obj <- RunUMAP(seurat_obj, dims = 1:10)
seurat_obj <- RunTSNE(seurat_obj, dims = 1:10)

seurat_obj <- SetupForWGCNA(
  seurat_obj,
  gene_select = "fraction"
  fraction = 0.05, 
  wgcna_name = "AOA"
)

seurat_obj <- MetacellsByGroups(
  seurat_obj = seurat_obj,
  group.by = c("seurat_clusters"),
  reduction = "pca",
  k = 20,
  min_cells = 20,
  max_shared = 10,
  ident.group = "seurat_clusters"
)

[smorabit]

Hi, I have not seen this particular error so I don't know the exact source of your problem. Not sure if this will work but can you please try to raise min_cells such that it is greater than k? Maybe just try the default settings for this function and see if it works.

[sandy627]

Hi，@smorabit，I also meet this trouble. This problem is also encountered with a large seruat object, but taking a subset of it works perfectly well. Large seurat object with about 100w cells.

[Li-ZhiD]

Same problem. @sandy627 downsampling solved the problem? Thank you!

[smorabit]

@sandy627
@Li-ZhiD

Can you please show me the code and error message?

[Li-ZhiD]

Setting ”min_cells = 100“ can solve this problem for my data.