Specificity control plot

Draws the arms of the specificity control side by side: the assay with the detection reagent present against the same exposure with the reagent omitted. No amount of batch standardisation can show that a readout is reagent-dependent rather than background autofluorescence, which is what makes this the strongest validation a method-establishment result has.

Usage

cr_plot_specificity(
  spec,
  arm = "arm",
  value = NULL,
  arm_levels = NULL,
  colour_by = NULL,
  log_y = TRUE,
  ratio_arms = NULL,
  title = NULL,
  subtitle = NULL,
  x_lab = NA,
  y_lab = NULL
)

Arguments

spec

Cell-level or per-arm data: a data frame, or a list carrying one in a spec or specificity element.

arm

Name of the column identifying the arm (default "arm").

value

Name of the value column. NULL (default) auto-detects: median_signal for a summarised table, otherwise value.

arm_levels

Optional character vector fixing the order of the arms.

colour_by

Column mapped to fill, or NULL to fill by arm.

log_y

Draw the signal axis on a log scale (default TRUE).

ratio_arms

Optional length-2 character vector naming the arms whose median ratio is reported in the subtitle, as c(numerator, denominator). When spec carries a signal_to_background attribute that value is used instead.

title, subtitle, x_lab, y_lab

Plot labels. NULL uses a computed default; NA omits the label.

Value

A ggplot object.

Details

Accepts either cell-level data (one row per cell, drawn as violins with the arm medians annotated) or an already-summarised table (one row per arm, drawn as columns).

See also

Other screen figures: cr_plot_forest(), cr_plot_qc_gate(), cr_plot_sample_size(), cr_plot_screen(), cr_save_plot()

Examples

set.seed(1)
spec <- data.frame(
  arm = rep(c("reagent omitted\n+ vehicle", "reagent omitted\n+ exposed",
              "reagent present\n+ vehicle", "reagent present\n+ exposed"),
            each = 60),
  value = c(rlnorm(60, 3, 0.3), rlnorm(60, 3.05, 0.3),
            rlnorm(60, 4.5, 0.3), rlnorm(60, 6.2, 0.3))
)
cr_plot_specificity(spec)