plot_hexbin_feature: Plot of feature expression of single cells in bivariate...
In schex: Hexbin plots for single cell omics data

Description Usage Arguments Details Value Examples

Plot of feature expression of single cells in bivariate hexagon cells.

plot_hexbin_feature(
  sce,
  mod = "RNA",
  type,
  feature,
  action,
  title = NULL,
  xlab = NULL,
  ylab = NULL,
  lower_cutoff = 0,
  upper_cutoff = 1
)

`sce`	A `SingleCellExperiment` or `Seurat-class` object.
`mod`	A string referring to the name of the modality used for plotting. For RNA modality use "RNA". For other modalities use name of alternative object for the `SingleCellExperiment` or the name of the assay for the `Seurat-class` object.
`type`	A string referring to the type of assay in the `SingleCellExperiment` object or the data transformation in the `Seurat-class` object.
`feature`	A string referring to the name of one feature.
`action`	A strings pecifying how meta data of observations in binned hexagon cells are to be summarized. Possible actions are `prop_0`, `mode`, `mean` and `median` (see details).
`title`	A string containing the title of the plot.
`xlab`	A string containing the title of the x axis.
`ylab`	A string containing the title of the y axis.
`lower_cutoff`	For `mode`, `mean` and `median` actions, remove expression values below this quantile. Expressed as decimal. Default: 0
`upper_cutoff`	For `mode`, `mean` and `median` actions, remove expression values above this quantile. Expressed as decimal. Default: 1

This function plots the expression of any feature in the hexagon cell representation calculated with make_hexbin. The chosen gene expression is summarized by one of four actions prop_0, mode, mean and median:

prop_0: Returns the proportion of observations in the bin greater than 0. The associated meta data column needs to be numeric.
mode: Returns the mode of the observations in the bin. The associated meta data column needs to be numeric.
mean: Returns the mean of the observations in the bin. The associated meta data column needs to be numeric.
median: Returns the median of the observations in the bin. The associated meta data column needs to be numeric.

To access the data that has been integrated in the Seurat-class object specifiy mod="integrated".

A ggplot2{ggplot} object.

# For Seurat object
library(Seurat)
data("pbmc_small")
pbmc_small <- make_hexbin(pbmc_small, 10, dimension_reduction = "PCA")
plot_hexbin_feature(pbmc_small, type="counts", feature="TALDO1", 
   action="median")
plot_hexbin_feature(pbmc_small, type="counts", feature="TALDO1", 
   action="median", lower_cutoff=0.2, upper_cutoff=0.5)
# For SingleCellExperiment object
## Not run: 
library(TENxPBMCData)
library(scater)
tenx_pbmc3k <- TENxPBMCData(dataset = "pbmc3k")
rm_ind <- calcAverage(tenx_pbmc3k)<0.1
tenx_pbmc3k <- tenx_pbmc3k[!rm_ind,]
colData(tenx_pbmc3k) <- cbind(colData(tenx_pbmc3k),
   perCellQCMetrics(tenx_pbmc3k))
tenx_pbmc3k <- logNormCounts(tenx_pbmc3k)
tenx_pbmc3k <- runPCA(tenx_pbmc3k)
tenx_pbmc3k <- make_hexbin( tenx_pbmc3k, 20, dimension_reduction = "PCA")
plot_hexbin_feature(tenx_pbmc3k, type="logcounts",
   feature="ENSG00000135250", action="median")
plot_hexbin_feature(tenx_pbmc3k, type="logcounts",
   feature="ENSG00000135250", action="mode")

## End(Not run)
# For other modalities in Seurat object
library(Seurat)
data("pbmc_small")
pbmc_small <- make_hexbin(pbmc_small, 10, dimension_reduction = "PCA")
protein <- matrix(rnorm(10* ncol(pbmc_small)), ncol=ncol(pbmc_small))
rownames(protein) <- paste0("A", seq(1,10,1))
colnames(protein) <- colnames(pbmc_small)
pbmc_small[["ADT"]] <- CreateAssayObject(counts = protein)
pbmc_small <- make_hexbin(pbmc_small, 10, dimension_reduction = "PCA")
plot_hexbin_feature(pbmc_small, type="counts", mod="ADT",
    feature="A1", action="prop_0")