Skip to contents

Compute per-cluster mean expression and expressing fraction for a gene set

compute_expression_info_sc.Rd

Thin R wrapper around the Rust compute_cluster_expression_stats routine. Streams gene chunks from the on-disk store and aggregates expression across user-supplied cell clusters. Cells outside any cluster are ignored.

If condition_colname and condition_oi are supplied, only cells from that condition contribute to the aggregation.

Usage

compute_expression_info_sc(
  object,
  celltype_colname,
  genes,
  condition_colname = NULL,
  condition_oi = NULL
)

Arguments

object

A SingleCells object.

celltype_colname

Name of the cluster column in obs.

genes

Character vector of gene IDs to aggregate over.

condition_colname

Optional. Name of a condition column in obs.

condition_oi

Optional. Value of condition_colname to subset to.

Value

A long data.table with columns cluster_id, gene, avg_expr, frac_expr.

Examples

# mean expression and expressing fraction per planted cell type
sc <- demo_single_cells()
res <- compute_expression_info_sc(
  sc,
  celltype_colname = "cell_grp",
  genes = get_gene_names(sc)[1:5]
)
head(res)
#>     cluster_id    gene avg_expr frac_expr
#>         <char>  <char>    <num>     <num>
#> 1: cell_type_1 gene_01 6.241650 0.9880240
#> 2: cell_type_1 gene_02 6.115141 0.9880240
#> 3: cell_type_1 gene_03 6.385783 1.0000000
#> 4: cell_type_1 gene_04 5.770888 0.9580838
#> 5: cell_type_1 gene_05 4.425770 0.8383234
#> 6: cell_type_2 gene_01 2.902391 0.6946108

unlink(sc@dir_data, recursive = TRUE, force = TRUE)