Skip to contents

Bubble plot of differential expression across treatments

Source: R/ptrap_bubble.R
ptrap_bubble.Rd

Displays genes of interest (y-axis) against treatment conditions (x-axis) as a bubble plot. Circle fill encodes the log2 fold change (logFC) via a continuous colour gradient, and circle size encodes statistical significance as \(-\log_{10}(\mathrm{FDR})\) (default) or \(-\log_{10}(p\text{-value})\).

Usage

ptrap_bubble(
  de_result,
  sig_size = c("FDR", "PValue"),
  colors_lfc = c("#25599b", "#ABD0DD", "#F2F9FE", "#F88705", "#B8351F"),
  gene_col = "Gene",
  treatment_col = "treatment",
  size_range = c(2, 12),
  point_alpha = 0.9,
  title = NULL,
  interactive = FALSE
)

Arguments

de_result

A data frame of combined results from one or more calls to ptrap_de(). For test_method = "paired.ttest", use the $results component of each and combine them with dplyr::bind_rows(). Must contain columns for gene names, treatment, logFC, FDR, and PValue.

sig_size

Character. Which column to use for the size aesthetic (displayed as \(-\log_{10}\)). Either "FDR" (default) or "PValue".

colors_lfc

Character vector of at least three colours defining the fill gradient mapped to logFC values. Passed to ggplot2::scale_fill_gradientn(). Default: c("#25599b", "#ABD0DD", "#F2F9FE", "#F88705", "#B8351F").

gene_col

Name of the column containing gene identifiers. Default "Gene".

treatment_col

Name of the column containing the treatment label. Default "treatment".

size_range

Numeric vector of length two controlling the minimum and maximum point sizes. Passed to ggplot2::scale_size_continuous(). Default c(2, 12).

point_alpha

Opacity of the bubbles. Default 0.9.

title

Optional plot title. If NULL (default), no title is added.

interactive

Logical. If TRUE, returns an interactive plotly::ggplotly() object with hover tooltips showing gene name, logFC, p-value, and FDR. Default FALSE.

Value

A ggplot2::ggplot() object, or a plotly object when interactive = TRUE.

Details

This is particularly useful after running ptrap_de() for multiple treatment conditions (e.g., PACAP and BDNF) on the same set of genes: bind the $results tibbles together with dplyr::bind_rows() and pass the combined data frame to ptrap_bubble().

Examples

if (FALSE) { # \dontrun{
# Run DE for two treatments on the same gene set
pacap_de <- ptrap_de(counts, ..., treatment_name = "PACAP",
                     genes.filter = my_genes)
bdnf_de  <- ptrap_de(counts, ..., treatment_name = "BDNF",
                     genes.filter = my_genes)

# Combine results and plot
library(dplyr)
bind_rows(pacap_de$results, bdnf_de$results) |>
  ptrap_bubble()

# Use raw p-values for size and a custom colour palette
bind_rows(pacap_de$results, bdnf_de$results) |>
  ptrap_bubble(sig_size = "PValue",
               colors_lfc = c("blue", "white", "red"))

# Interactive with hover tooltips
bind_rows(pacap_de$results, bdnf_de$results) |>
  ptrap_bubble(interactive = TRUE)
} # }