Run CopyKAT copy-number inference

Fast, reference-optional CNV calling using the CopyKAT Bayesian model. Useful as a quick first-pass or to validate RunInferCNV() results. Adds copykat_prediction ("aneuploid" / "diploid" / "not.defined") and copykat_score to Seurat metadata.

Usage

RunCopyKAT(
  seurat_object,
  cell_type_col = "Assignment",
  ref_cells = NULL,
  output_dir = NULL,
  n_cores = 4L,
  LOW_DR = 0.05,
  UP_DR = 0.1,
  win.size = 25L,
  caffeinate = FALSE
)

Arguments

seurat_object

A Seurat object.

cell_type_col

Character. Metadata column used to color output plots. Default "Assignment".

ref_cells

Character vector of cell-type labels to use as the diploid reference. NULL runs in unsupervised mode (CopyKAT infers normals automatically). Default NULL.

output_dir

Character or NULL. Directory for CopyKAT output files and plots. Walks up from PrepObject() when NULL.

n_cores

Integer. Parallel threads. Default 4L.

LOW_DR, UP_DR

Numeric. Lower and upper bounds on the dropout rate filter passed to CopyKAT. Defaults 0.05 / 0.1.

win.size

Integer. Genomic window size (number of genes) for CNV smoothing. Default 25.

caffeinate

Logical. Prevent the machine from sleeping. Default FALSE.

Value

The input Seurat object with copykat_prediction and copykat_score added to metadata.

See also

RunInferCNV()