Run MAD outlier detection on per-cell QC metrics — run_cell_qc • bixverse

Run MAD outlier detection on per-cell QC metrics

Usage

run_cell_qc(metrics, directions = NULL, threshold = 3)

Arguments

metrics: Named list of numeric vectors. Each element is a QC metric to check (e.g. list(log10_lib_size = log10(lib_size), MT = mt_pct)).
directions: Named character vector mapping metric names to direction. One of "twosided", "below", "above". Defaults to "twosided" for all metrics if NULL.
threshold: Numeric. Number of MADs to use for outlier detection.

Value

An object of class CellQc containing:

metrics: The input metrics list.
per_metric: Named list of per-metric results from per_cell_qc_outlier.
outlier_mat: Logical matrix with one column per metric.
combined: Logical vector. TRUE if a cell is an outlier in any metric.