Plot top PC loadings of genes

Just like Seurat's VizDimLoadings function. I haven't found an equivalent for SCE but find it useful. But I'm not trying to reproduce that Seurat function exactly. For instance, I don't like it when Seurat imposes a ggplot theme, and I don't like the cowplot theme. Maybe I should rewrite it in base R but for now I'm using Tidyverse.

Usage

plotDimLoadings(
  sce,
  dims = 1:4,
  nfeatures = 10,
  swap_rownames = NULL,
  reduction = "PCA",
  balanced = TRUE,
  ncol = 2,
  sample_id = "all"
)

Arguments

sce

A SingleCellExperiment object, or anything that inherits from SingleCellExperiment.

dims

Numeric vector specifying which PCs to plot. For MULTISPATI, PCs with negative eigenvalues are in the right most columns of the embedding and loading matrices. See the ElbowPlot.

nfeatures

Number of genes to plot.

swap_rownames

Column name of rowData(object) to be used to identify features instead of rownames(object) when labeling plot elements. If not found in rowData, then rownames of the gene count matrix will be used.

reduction

Name of the dimension reduction to use. It must have an attribute called either "percentVar" or "eig" for eigenvalues. Defaults to "PCA".

balanced

Return an equal number of genes with + and - scores. If FALSE, returns the top genes ranked by the scores absolute values.

ncol

Number of columns in the facetted plot.

sample_id

Sample(s) in the SFE object whose cells/spots to use. Can be "all" to compute metric for all samples; the metric is computed separately for each sample.

Value

A ggplot object. Loadings for different PCs are plotted in different facets so one ggplot object is returned.

Examples

library(SFEData)
library(scater)
sfe <- McKellarMuscleData("small")
#> see ?SFEData and browseVignettes('SFEData') for documentation
#> loading from cache
sfe <- runPCA(sfe, ncomponents = 10, exprs_values = "counts")
plotDimLoadings(sfe, dims = 1:2)