Create .cells Object with Automatic Format Detection

Convenience wrapper that automatically detects input format and calls the appropriate conversion function. Supports Seurat (v4/v5), SingleCellExperiment, and list of count matrices. Simplifies workflow by eliminating need to know which specific function to call.

Usage

get.cells(
  .exprs,
  .meta,
  .sample.var = NULL,
  .assay = "RNA",
  .layer.pattern = "counts",
  .slot = "counts",
  .min.cells.per.sample = 10,
  .compress = FALSE,
  .verbose = TRUE
)

Arguments

.exprs

Expression data in one of three formats:

Named list of count matrices (calls get.cells.list.mat)
Seurat object v4 or v5 (calls get.cells.Seurat or get.cells.Seurat5)
SingleCellExperiment object (calls get.cells.SCE)

.meta

Data frame with sample-level metadata. Required for list input. Rownames must match sample IDs. For Seurat/SCE, rownames must match values in .sample.var column.

.sample.var

Character: metadata column identifying samples. Required for Seurat/SCE, ignored for list input.

.assay

Character: assay name. Default "RNA" for Seurat, "counts" for SCE. Ignored for list input.

.layer.pattern

Character: layer name pattern for Seurat v5. Default "counts". Ignored for other formats.

.slot

Character: data slot for Seurat v4. Default "counts". Ignored for other formats.

.min.cells.per.sample

Integer: minimum cells per sample. Default 10. Ignored for list input.

.compress

Logical: compress RDS files? Default FALSE.

.verbose

Logical: print progress messages? Default TRUE.

Value

Named list of file paths to temporary RDS files, one per sample.

Details

This function inspects the class of .exprs to determine format:

Seurat: Checks for Assay5 class to distinguish v4 from v5
SingleCellExperiment: Maps "RNA" assay to "counts" if needed
list: Validates names match .meta rownames

For Seurat v5, automatically handles layered data structure. For SCE, uses standard assay() accessor.

Examples

if (FALSE) { # \dontrun{
# Example 1: List of count matrices
count.matrices <- list(A_R1 = counts.A_R1, B_R1 = counts.B_R1)
cells <- get.cells(.exprs = count.matrices,
                   .meta = sim_trajectory.meta[c("A_R1", "B_R1"), ])

# Example 2: Seurat object (auto-detects v4 vs v5)
cells <- get.cells(.exprs = seurat.obj,
                   .meta = sim_trajectory.meta,
                   .sample.var = "sample_id")

# Example 3: SingleCellExperiment object
cells <- get.cells(.exprs = sce.obj,
                   .meta = sim_trajectory.meta,
                   .sample.var = "Sample",
                   .assay = "counts")
} # }