Skip to contents

This is a transparent screening heuristic, not a test that automatically excludes samples. Distances are calculated from median/MAD-standardized PC scores and compared with a chi-squared quantile.

Usage

qc_outliers(
  pca,
  components = seq_len(min(5L, ncol(pca$x))),
  probability = 0.99
)

Arguments

pca

A prcomp object, usually from reduce_pca().

components

Principal components to include.

probability

Chi-squared cutoff probability.

Value

A data frame containing robust distance, cutoff, and flag.