Introduction to scmisc

Installation

You can install the release version of scmisc from Github with:

remotes::install_github("ddiez/scmisc")

Toy dataset

This toy dataset contains 493 cells from mice in a SingleCellExperiment object.

sce
#> class: SingleCellExperiment 
#> dim: 434 493 
#> metadata(0):
#> assays(2): counts logcounts
#> rownames(434): X2810417H13Rik Ada ... Zfp36l2 Zfp36
#> rowData names(4): id symbol accession note
#> colnames(493): S00001_274363 S00001_323404 ... S00001_843408
#>   S00001_270588
#> colData names(3): samplename filename cell_index
#> reducedDimNames(0):
#> mainExpName: NULL
#> altExpNames(0):

We can create some grouping variables.

sce$group1 <- sample(c("A", "B"), ncol(sce), replace = TRUE)
sce$group2 <- sample(c("C", "D"), ncol(sce), replace = TRUE)

Preprocess

To calculate PCA we can use reduce_dim.

sce <- reduce_dim(sce, method = "pca")

Plots

There are several plotting functions that work with Seurat and SingleCellExperiment objects.

Scatter plot

plot_coord(sce)

plot_coord(sce, color = "group1")

plot_coord(sce, color = "group2")

Expand by grouping variable.

plot_coord(sce, expand = "group1")


plot_coord(sce, expand = c("group1", "group2"))

Heatmap

The function plot_heatmap is a wrapper around ComplexHeatmap::Heatmap that simplifies plotting SingleCellExperiment and Seurat objects.

plot_heatmap(sce[1:20, ])

We can use ComplexHeatmap advance annotation features with the heatmap by passing them directly.

cols <- circlize::colorRamp2(c(-2, 0, 2), c("purple", "black", "yellow"))
plot_heatmap(sce[1:20, ], col = cols)

plot_heatmap(sce[1:20, ], column_split = sce$group1)

For column annotations we can directly pass the variable names to use with the color specification using top_ann and top_ann_col.

col <- list(group1 = c(A = "lightgrey", B = "black"),
            group2 = c("C" = "limegreen", "D" = "violetred"))
plot_heatmap(sce[1:20, ], column_split = sce$group1, top_ann = c("group1", "group2"), top_ann_col = col)