Skip to contents

Principal component regression on batch

rs_pcr.Rd

[Experimental] Regresses each embedding dimension on the batch labels (one-way ANOVA) and weights the per-dimension R-squared by the variance each dimension carries. Compare the value on the uncorrected PCA with the one on the corrected embedding, (pre - post) / pre, rather than reading it on its own.

Usage

rs_pcr(embedding, batch_vector)

Arguments

embedding

Numeric matrix. Cells x dimensions, ideally a PCA.

batch_vector

Integer vector. The batch per cell. The codes need not be 0-based or contiguous.

Value

A list with the following items

  • var_explained - Variance per embedding dimension.

  • r_squared - R-squared of batch per embedding dimension.

  • pcr - Variance-weighted R-squared of batch.

References

Büttner, et al., Nat Methods, 2019; Luecken, et al., Nat Methods, 2022