R/plot_gene_express.R

In lahuuki/DeconvoBuddies: Helper Functions for LIBD Deconvolution

#' Plot the gene expression of a list of genes in a SCE object
#'
#' This function plots the expression of one or more genes as a violin plot,
#' over a user defined category, typically a cell type annotation. The plots are
#' made using `ggplot2`.
#'
#' @param sce A
#' [SummarizedExperiment-class][SummarizedExperiment::SummarizedExperiment-class]
#' object or one inheriting it.
#' @param genes  A `character()` vector specifying the genes to plot,
#' this should match the format of `rownames(sce)`.
#' @param assay_name A `character(1)` specifying the name of the
#' [assay()][SummarizedExperiment::SummarizedExperiment-class] in the
#' `sce` object to use to rank expression values. Defaults to `logcounts` since
#' it typically contains the normalized expression values.
#' @param category  A `character(1)` specifying the name of the categorical
#' variable to group the cells or nuclei by. Defaults to `cellType`.
#' @param color_pal  A named `character(1)` vector that contains a color pallet
#' matching the `category` values.
#' @param title A `character(1)` to title the plot.
#' @param plot_points A `logical(1)` indicating whether to plot points over the
#' violin, defaults to `FALSE` as these often become over plotted and quite
#' large (especially when saved as PDF).
#' @param ncol An `integer(1)` specifying the number of columns for the facet in
#' the final plot. Defaults to 2.
#'
#' @return A `ggplot()` violin plot for selected genes.
#' @export
#'
#' @examples
#' ## Using Symbol as rownames makes this more human readable
#' data(sce_ab)
#' plot_gene_express(sce = sce_ab, genes = c("G-D1_A"))
#'
#' # Access example data
#' if (!exists("sce_DLPFC_example")) sce_DLPFC_example <- fetch_deconvo_data("sce_DLPFC_example")
#'
#' ## plot expression of two genes
#' plot_gene_express(
#'     sce = sce_DLPFC_example,
#'     category = "cellType_broad_hc",
#'     genes = c("GAD2", "CD22")
#' )
#'
#' ## plot points - note this creates large images and is easy to over plot
#' plot_gene_express(
#'     sce = sce_DLPFC_example,
#'     category = "cellType_broad_hc",
#'     genes = c("GAD2", "CD22"), plot_points = TRUE
#' )
#'
#' ## Add title
#' plot_gene_express(
#'     sce = sce_DLPFC_example,
#'     category = "cellType_broad_hc",
#'     genes = c("GAD2", "CD22"),
#'     title = "My Genes"
#' )
#'
#' @family expression plotting functions
#'
plot_gene_express <- function(
        sce,
        genes,
        assay_name = "logcounts",
        category = "cellType",
        color_pal = NULL,
        title = NULL,
        plot_points = FALSE,
        ncol = 2) {
    stopifnot(any(genes %in% rownames(sce)))

    if (!category %in% colnames(colData(sce))) {
        stop(
            category,
            "' is not a column name in colData(sce), check that `category` matches this sce",
            call. = FALSE
        )
    }

    stopifnot(assay_name %in% SummarizedExperiment::assayNames(sce))

    value <- median <- NULL

    category_df <- as.data.frame(colData(sce))[, category, drop = FALSE]
    expression_long <- reshape2::melt(as.matrix(SummarizedExperiment::assays(sce)[[assay_name]][genes, , drop = FALSE]))

    category <- category_df[expression_long$Var2, ]
    expression_long <- cbind(expression_long, category)

    expression_violin <- ggplot(data = expression_long, aes(x = category, y = value)) +
        # ggplot2::geom_violin(aes(fill = category), scale = "width") +
        ggplot2::facet_wrap(~Var1, ncol = ncol) +
        ggplot2::labs(
            y = paste0("Expression (", assay_name, ")"),
            title = title
        ) +
        ggplot2::theme_bw() +
        ggplot2::theme(
            legend.position = "None",
            axis.title.x = ggplot2::element_blank(),
            axis.text.x = ggplot2::element_text(angle = 90, hjust = 1),
            strip.text.x = ggplot2::element_text(face = "italic")
        )

    if (plot_points) {
        expression_violin <- expression_violin +
            ggplot2::geom_violin(aes(color = category), scale = "width") +
            ggplot2::geom_jitter(aes(color = category),
                position = ggplot2::position_jitter(seed = 1, width = 0.2), size = .5
            ) +
            ggplot2::stat_summary(
                fun = mean,
                geom = "crossbar",
                width = 0.3
            )

        if (!is.null(color_pal)) expression_violin <- expression_violin + ggplot2::scale_color_manual(values = color_pal)

        return(expression_violin)
    }

    expression_violin <- expression_violin +
        ggplot2::geom_violin(aes(fill = category), scale = "width") +
        ggplot2::stat_summary(
            fun = mean,
            geom = "crossbar",
            width = 0.3
        )

    if (!is.null(color_pal)) expression_violin <- expression_violin + ggplot2::scale_fill_manual(values = color_pal)

    # expression_violin
    return(expression_violin)
}