The goal of schex is to provide easy plotting of hexagon cell representations of single cell data stored in SingleCellExperiment objects.

You can install schex using the Bioconductor project:

# install.packages("BiocManager")
BiocManager::install("schex")

You can install the development version of schex with:

# install.packages("devtools")
devtools::install_github("SaskiaFreytag/schex")

Did you know that the order in which points are plotted depends on their location in the data frame? For example when plotting the expression of CD19, a B-cell maker, you may get the following three plots depending on how you order your observations.

Using the first plot you would not decide to call the central cluster a B-cell population. Using the second plot you would probably decide to call the same cluster a B-cell population. Using the last plot you might be undecided.

Instead of plotting points on top of each other, schex summarizes points into hexagon cells. Hence avoiding confusion due to observation order.

Check out the vignettes to learn about how to get started. Or for a quick start, use the following code.

library(TENxPBMCData)
library(scater)
tenx_pbmc3k <- TENxPBMCData(dataset = "pbmc3k")
rm_ind <- calculateAverage(tenx_pbmc3k) < 0.1
tenx_pbmc3k <- tenx_pbmc3k[!rm_ind, ]
tenx_pbmc3k <- logNormCounts(tenx_pbmc3k)
tenx_pbmc3k <- runPCA(tenx_pbmc3k)
tenx_pbmc3k <- make_hexbin(tenx_pbmc3k, 10, dimension_reduction = "PCA")
plot_hexbin_density(tenx_pbmc3k)