Filter cells in a Seurat object based on QC metrics

This function filters cells in a Seurat object using Median Absolute Deviation (MAD) to identify outliers across specified metadata features. Supports grouped analysis and provides visual diagnostics.

Usage

sn_filter_cells(
  x,
  features,
  group_by = NULL,
  method = "mad",
  n = 5,
  plot = TRUE,
  filter = TRUE
)

Arguments

x: A Seurat object
features: Character vector of metadata column names to use for filtering
group_by: (Optional) Metadata column name to group by for group-wise calculations
method: Outlier detection method (currently only "mad" supported)
n: Numeric threshold(s) for MAD multiplier. Can be single value or vector matching features.
plot: Logical indicating whether to generate QC diagnostic plots
filter: Logical indicating whether to filter out flagged cells

Value

Seurat object with QC flags in metadata. If filter=TRUE, returns subsetted object.

Examples

if (FALSE) { # \dontrun{
# Basic usage
pbmc <- sn_filter_cells(pbmc, features = c("nCount_RNA", "nFeature_RNA"))

# Grouped analysis
pbmc <- sn_filter_cells(pbmc, features = "percent.mt", group_by = "sample")

# Custom threshold with plotting
pbmc <- sn_filter_cells(pbmc, features = "nCount_RNA", n = 3, plot = TRUE)
} # }