Aggregated feature expression by identity class — AggregateExpression • Seurat

Returns summed counts ("pseudobulk") for each identity class.

AggregateExpression(
  object,
  assays = NULL,
  features = NULL,
  return.seurat = FALSE,
  group.by = "ident",
  add.ident = NULL,
  normalization.method = "LogNormalize",
  scale.factor = 10000,
  margin = 1,
  verbose = TRUE,
  ...
)

Arguments

object: Seurat object
assays: Which assays to use. Default is all assays
features: Features to analyze. Default is all features in the assay
return.seurat: Whether to return the data as a Seurat object. Default is FALSE
group.by: Category (or vector of categories) for grouping (e.g, ident, replicate, celltype); 'ident' by default To use multiple categories, specify a vector, such as c('ident', 'replicate', 'celltype')
add.ident: (Deprecated). Place an additional label on each cell prior to pseudobulking
normalization.method: Method for normalization, see NormalizeData
scale.factor: Scale factor for normalization, see NormalizeData
margin: Margin to perform CLR normalization, see NormalizeData
verbose: Print messages and show progress bar
...: Arguments to be passed to methods such as CreateSeuratObject

Value

Returns a matrix with genes as rows, identity classes as columns. If return.seurat is TRUE, returns an object of class Seurat.

Details

If return.seurat = TRUE, aggregated values are placed in the 'counts' layer of the returned object. The data is then normalized by running NormalizeData on the aggregated counts. ScaleData is then run on the default assay before returning the object.

Examples

if (FALSE) {
data("pbmc_small")
head(AggregateExpression(object = pbmc_small)$RNA)
head(AggregateExpression(object = pbmc_small, group.by = c('ident', 'groups'))$RNA)
}