Generate SEACells — rs_get_seacells • bixverse

This function implements the SEACells algorithm for generating meta cells from Persad et al. An embedding matrix must be provided which is used to construct the kNN graph and kernel matrix for the SEACells algorithm. This version is highly memory and speed-optimised and will truncate small values during matrix operations which can affect convergence.

Usage

rs_get_seacells(
  f_path,
  embd,
  cells_to_keep,
  cells_to_use,
  seacells_params,
  target_size,
  seed,
  verbose
)

Arguments

f_path: String. Path to the counts_cells.bin file.
embd: Numerical matrix. The embedding matrix (for example PCA embedding) used for the generation of the kNN graph and kernel matrix.
cells_to_keep: Optional indices of the cells to keep, i.e., the cells used for the generation of the embedding.
cells_to_use: Optional indices of cells to use for meta cell generation. Useful if you wish to generate meta cells in specific cell types.
seacells_params: A list containing the SEACells parameters.
target_size: Numeric. Target library size for re-normalisation of the meta cells. Typically 1e4.
seed: Integer. For reproducibility purposes.
verbose: Boolean. Controls verbosity of the function.

Value

A list with the following elements:

assignments - A list containing assignment information with elements: assignments (vector), metacells (list), unassigned (vector), n_metacells, n_cells, n_unassigned
aggregated - A list with indptr, indices, raw_counts, norm_counts, nrow, ncol in sparse format.
rss - Vector of RSS values from each iteration.
archetypes - Vector of cell indices selected as archetypes.

References

Persad, et al., Nat. Biotechnol., 2023.