Key characteristics comparison

Overview

This script shows one example PBMC10k: CD8 effector cell type to generate figure 2 in the manuscript.
It also shows supplementary plots associated with figure 2.

Same script can be used to generate plots for other datasets or by directly using simPICcompare function

Load data

Loading example data form PBMC10k

Show code

p <- readRDS("../data/Figure2/pbmc10k_cd8effector.rds")

Setting plot aesthetics

Show code

features <- p$Plots$Means$data
cells <- p$Plots$LibrarySizes$data

point.size = 0.2
point.alpha = 0.1

colours <- c(
  "#E41A1C", "#377EB8", "#4DAF4A", "#984EA3",
  "#008080", "#FFA07A", "#7B68EE", "#9ACD32", "black"
)

Show code

plot_theme <- function() {
  ggplot2::theme(
    axis.text = ggplot2::element_text(size = 12, colour = "black"),
    axis.title = ggplot2::element_text(size = 12),
    panel.grid = ggplot2::element_blank(),
    panel.border = ggplot2::element_rect(
      color = "black", fill = NA,
      size = 1
    ),
    panel.background = ggplot2::element_blank(),
    plot.background = ggplot2::element_blank(),
    plot.title = ggplot2::element_text(face = "bold", color = "black"),
    legend.position = "none"
  )
}


libs <- ggplot2::ggplot(
  cells,
  ggplot2::aes(
    x = .data$Dataset, y = .data$sum, colour = .data$Dataset,
    fill = .data$Dataset
  )
) +
  ggplot2::geom_boxplot(
    width = 0.8, size = 0.6, alpha = 0.3,
    position = ggplot2::position_dodge2(0.5)
  ) +
  ggplot2::scale_y_continuous(labels = scales::comma) +
  ggplot2::scale_colour_manual(values = colours) +
  ggplot2::scale_fill_manual(values = colours) +
  ggplot2::xlab("") +
  ggplot2::ylab("library size") +
  ggplot2::ggtitle("Distribution of library sizes") +
  plot_theme()

z.cell <- ggplot2::ggplot(
  cells,
  ggplot2::aes(
    x = .data$Dataset, y = .data$PctZero,
    colour = .data$Dataset, fill = .data$Dataset
  )
) +
  ggplot2::geom_boxplot(
    width = 0.8, size = 0.6, alpha = 0.3,
    position = ggplot2::position_dodge2(0.5)
  ) +
  ggplot2::ylim(50, quantile(features$PctZero, 1)) +
  ggplot2::scale_colour_manual(values = colours) +
  ggplot2::scale_fill_manual(values = colours) +
  ggplot2::xlab("") +
  ggplot2::ylab("% zeros per cell") +
  ggplot2::ggtitle("Distribution of zeros per cell") +
  plot_theme()


means <- ggplot2::ggplot(
        features,
        ggplot2::aes(
            x = .data$Dataset, y = .data$mean,
            colour = .data$Dataset, fill = .data$Dataset
        )
    ) +
        ggplot2::geom_boxplot(
            width = 0.8, size = 0.6, alpha = 0.3,
            position = ggplot2::position_dodge2(0.5)
        ) +
        ggplot2::scale_colour_manual(values = colours) +
        ggplot2::scale_fill_manual(values = colours) +
        ggplot2::xlab("") +
        ggplot2::ylab("Peak means") +
        ggplot2::ggtitle("Distribution of peak means") +
        plot_theme()

vars <- ggplot2::ggplot(
  features,
  ggplot2::aes(
    x = .data$Dataset, y = .data$VarCounts,
    colour = .data$Dataset, fill = .data$Dataset
  )
) +
  ggplot2::geom_boxplot(
    width = 0.8, size = 0.6, alpha = 0.3,
    position = ggplot2::position_dodge2(0.5)
  ) +
  ggplot2::scale_colour_manual(values = colours) +
  ggplot2::scale_fill_manual(values = colours) +
  ggplot2::xlab("") +
  ggplot2::ylab("Variance of peaks") +
  ggplot2::ggtitle("Distribution of peak variances") +
  plot_theme()

mean.var <- ggplot2::ggplot(
  features,
  ggplot2::aes(
    x = .data$mean, y = .data$VarCounts,
    fill = .data$Dataset, color = .data$Dataset
  )
) +
  ggplot2::geom_point(size = point.size, alpha = point.alpha) +
  ggplot2::scale_colour_manual(values = colours) +
  ggplot2::scale_fill_manual(values = colours) +
  ggplot2::xlab("Peak-Means") +
  ggplot2::ylab("Peak Variance") +
  ggplot2::ggtitle("Mean-variance relationship") +
  plot_theme()

z.peak <- ggplot2::ggplot(
  features,
  ggplot2::aes(
    x = .data$Dataset, y = .data$PctZero,
    color = .data$Dataset, fill = .data$Dataset
  )
) +
  ggplot2::geom_boxplot(
    width = 0.8, size = 0.6, alpha = 0.3,
    position = ggplot2::position_dodge2(0.5)
  ) +
  ggplot2::scale_y_continuous(limits = c(0, 100)) +
  ggplot2::scale_colour_manual(values = colours) +
  ggplot2::scale_fill_manual(values = colours) +
  ggplot2::xlab("") +
  ggplot2::ylab("% zeros per peak") +
  ggplot2::ggtitle("Distribution of zeros per peak") +
  plot_theme()

mean.zeros <- ggplot2::ggplot(
  features,
  ggplot2::aes(
    x = .data$mean, y = .data$PctZero,
    color = .data$Dataset, fill = .data$Dataset
  )
) +
  ggplot2::geom_point(size = point.size, alpha = point.alpha) +
  ggplot2::scale_x_log10(labels = scales::comma) +
  ggplot2::scale_colour_manual(values = colours) +
  ggplot2::scale_fill_manual(values = colours) +
  ggplot2::xlab("Peak-Means") +
  ggplot2::ylab("% zeros") +
  ggplot2::ggtitle("Mean-zeros relationship") +
  plot_theme()

pm.nzp <- ggplot2::ggplot(
  features,
  ggplot2::aes(
    x = .data$nzp, y = .data$mean,
    fill = .data$Dataset, color = .data$Dataset
  )
) +
  ggplot2::geom_point(size = point.size, alpha = point.alpha) +
  ggplot2::scale_colour_manual(values = colours) +
  ggplot2::scale_fill_manual(values = colours) +
  ggplot2::xlab("Non-zero proportion") +
  ggplot2::ylab("Peak-Means") +
  ggplot2::ggtitle("Correlation- peak mean and non-zero proportion") +
  plot_theme()

Generating Figure 2

First row is used as main figure 2

Show code

Plots = list( LibrarySizes =
libs, Means = means, ZerosCell = z.cell, Variances = vars, MeanVar =
mean.var, ZerosPeak = z.peak, MeanZeros = mean.zeros, PeakMeanNzp =
pm.nzp )

wrap_plots(Plots)