Pre-scan multiple h5ad files for multi-sample loading

Usage

prescan_h5ad_files(
  h5_paths,
  gene_universe = c("intersection", "union"),
  var_index = "_index",
  raw_count_slot = c("auto", "X", "raw.X", "layers.counts"),
  .verbose = TRUE
)

Arguments

h5_paths: Character vector of file paths to h5ad files. If names are provided, these will be used as experimental identifiers.
gene_universe: One of "intersection" or "union".
var_index: String. The name within the h5ad var part in which the variable names are stored. Defaults to "_index".
raw_count_slot: Where raw counts live. "auto" detects per file via detect_raw_count_slot(); otherwise one of "X", "raw.X", "layers.counts".
.verbose: Boolean. Controls verbosity of the function.

Value

A list with:

universe - Character vector of gene names in the universe
file_tasks - List of per-file task structures, each containing: exp_id, h5_path, cs_type, no_cells, no_genes, gene_local_to_universe (integer vector, NA for genes not in universe, 0-indexed into universe)

Examples

# build the gene universe across two files before a multi-file load
data <- generate_single_cell_test_data(
  syn_data_params = params_sc_synthetic_data(n_cells = 200L, n_genes = 40L)
)
files <- c(a = tempfile(fileext = ".h5ad"), b = tempfile(fileext = ".h5ad"))
for (f in files) {
  write_h5ad_sc(f, data$counts, data$obs, data$var, .verbose = FALSE)
}
tasks <- prescan_h5ad_files(h5_paths = files, .verbose = FALSE)
tasks$universe_size
#> [1] 40

unlink(files)