TestPCA: Test PCA

In Ghosn-Lab/BatchNorm: Data Matrix Normalization and Merging Strategies Minimize Batch-specific Systemic Variation in scRNA-Seq Data

Description

Identify important PCs from total principal components analysis (within a Seurat object). TestPCA functions by generating a z-score corresponding to each respective PC's proportional contribution to the total variance. Can be used similarly to the Seurat function ElbowPlot, which plots each successive PC by its standard deviation.

Usage

TestPCA(
  object = NULL,
  genes.use = object@assays$[e-mail],
  mtx.use = object@assays$[e-mail]
)

Arguments

object	A seurat object with variable features set and data scaled
genes.use	(Optional) The vector of variable features used to construct the PCs. Can be used in place of "object" to apply a specific set of variable genes.
mtx.use	(Optional) The expression matrix used to construct the PCs. Can be used in place of "object" to apply a specific matrix (such as non-scaled data or ADT counts). At a minimum, this matrix must included the variable features included in "genes.use"

Value

Returns a table containing the z-score of the cumulative percent of total variance for each PC

Examples

TestPCA(obj = PBMC4A)[, 1:5]

Ghosn-Lab/BatchNorm documentation built on Dec. 17, 2021, 9:32 p.m.