Skip to contents

Dual volcano plot comparing two treatment conditions from TRAP-seq DE results

Source: R/ptrap_volcano2.R
ptrap_volcano2.Rd

Takes two tibbles produced by ptrap_de() — one per treatment condition, both from the same brain region — joins them by gene, classifies each gene according to its differential expression status in each condition, and returns a scatter plot of logFC\(_{\text{treatment 1}}\) vs logFC\(_{\text{treatment 2}}\). Significant genes are highlighted and labelled; threshold lines are drawn at ±lfc_threshold on both axes.

Usage

ptrap_volcano2(
  de_result_1,
  de_result_2,
  fdr = TRUE,
  lfc_threshold = 1,
  fdr_threshold = 0.05,
  gene_col = "Gene",
  treatment_col = "Treatment",
  region_col = "BrainRegion",
  colors = NULL,
  point_size = 3.5,
  point_alpha = 0.7,
  genes.annot = NULL,
  max_overlaps = 20,
  title = NULL
)

Arguments

de_result_1

A tibble returned by ptrap_de() for the first treatment condition (plotted on the y-axis).

de_result_2

A tibble returned by ptrap_de() for the second treatment condition (plotted on the x-axis).

fdr

Logical. If TRUE (default), significance is assessed using the FDR column (BH-adjusted p-values). If FALSE, the raw PValue column is used instead.

lfc_threshold

Minimum absolute log2 fold change required to classify a gene as differentially expressed in a given condition (also sets the vertical and horizontal threshold lines). Default is 1.

fdr_threshold

P-value cutoff used to classify genes as DE. Applied to FDR when fdr = TRUE and to PValue when fdr = FALSE. Default is 0.05.

gene_col

Name of the column containing gene identifiers in both result tibbles. Default is "Gene".

treatment_col

Name of the column containing the treatment label in both result tibbles. Used to auto-generate axis labels and DE class names. Default is "Treatment".

region_col

Name of the column containing the brain region label. Used to set the default plot title. Default is "BrainRegion".

colors

A named character vector mapping each DE class to a colour. Names must be "DE in both", "DE only <t1>", and "DE only <t2>", where <t1> / <t2> are the treatment names found in the data. If NULL (default), a colourblind-friendly palette is used.

point_size

Size of the points. Default is 3.5.

point_alpha

Opacity of the coloured (significant) points. Default is 0.7.

genes.annot

Character vector of gene names to label on the plot via ggrepel::geom_text_repel(). Names must match values in the column specified by gene_col; an error is raised if any names are not found. Default is NULL (no labels).

max_overlaps

Passed to ggrepel::geom_text_repel(). Controls how many overlapping labels are allowed before they are hidden. Default is 20.

title

Plot title. If NULL (default), the brain region name extracted from de_result_1 is used.

Value

A ggplot2::ggplot() object.

Details

Significance can be assessed using either FDR-adjusted p-values (fdr = TRUE, default) or raw p-values (fdr = FALSE). In both cases the cutoff is fdr_threshold and the classification is computed inside this function from the supplied thresholds.

Examples

if (FALSE) { # \dontrun{
res_pb  <- ptrap_de(counts_mat, sample_df, gene_ids,
                    region_name = "POA", treatment_name = "pb")
res_sol <- ptrap_de(counts_mat, sample_df, gene_ids,
                    region_name = "POA", treatment_name = "sol")

# Default plot — FDR on both axes
ptrap_volcano2(res_pb, res_sol)

# Use raw p-values for the DE classification
ptrap_volcano2(res_pb, res_sol, fdr = FALSE)

# Custom thresholds, title and colours
ptrap_volcano2(res_pb, res_sol,
               lfc_threshold = 0.5,
               fdr_threshold = 0.1,
               title         = "POA: pb vs sol",
               colors        = c("DE in both"  = "#D55E00",
                                 "DE only pb"  = "#0072B2",
                                 "DE only sol" = "#49a15f"))
} # }