Make a "mushroom plot" of ligand-receptor interactions

make_mushroom_plot Make a plot in which each glyph consists of two semicircles corresponding to ligand- and receptor- information. The size of the semicircle is the percentage of cells that express the protein, while the saturation corresponds to the scaled average expression value.

Usage

make_mushroom_plot(
  prioritization_table,
  top_n = 30,
  show_rankings = FALSE,
  show_all_datapoints = FALSE,
  true_color_range = TRUE,
  use_absolute_rank = FALSE,
  size = "scaled_avg_exprs",
  color = "scaled_p_val_adapted",
  ligand_fill_colors = c("#DEEBF7", "#08306B"),
  receptor_fill_colors = c("#FEE0D2", "#A50F15"),
  unranked_ligand_fill_colors = c(alpha("#FFFFFF", alpha = 0.2), alpha("#252525", alpha =
    0.2)),
  unranked_receptor_fill_colors = c(alpha("#FFFFFF", alpha = 0.2), alpha("#252525", alpha
    = 0.2)),
  label_size = 3.88,
  ...
)

Arguments

prioritization_table

A prioritization table as generated by generate_prioritization_tables.

top_n

An integer indicating how many ligand-receptor pairs to show

show_rankings

A logical indicating whether to show the ranking of the ligand-receptor pairs (default: FALSE)

show_all_datapoints

A logical indicating whether to show all ligand-receptor pairs (default: FALSE, if true they will be grayed out)

true_color_range

A logical indicating whether to use the default color range as determined by ggplot (TRUE, default) or set the limits to a range of 0-1 (FALSE)

use_absolute_rank

A logical indicating to whether use the absolute prioritization rank to filter the top_n ligand-receptor pairs (default: FALSE)

size

A string indicating which column to use for the size of the semicircles (default: "scaled_avg_exprs"; use column name without "_ligand" or "_receptor" suffix)

color

A string indicating which column to use for the color of the semicircles (default: "scaled_p_val_adapted"; use column name without "_ligand" or "_receptor" suffix)

ligand_fill_colors

A vector of the low and high colors to use for the ligand semicircle fill gradient (default: c("#DEEBF7", "#08306B"))

receptor_fill_colors

A vector of the low and high colors to use for the receptor semicircle fill gradient (default: c("#FEE0D2", "#A50F15"))

unranked_ligand_fill_colors

A vector of the low and high colors to use for the unranked ligands when show_all_datapoints is TRUE (default: c(alpha("#FFFFFF", alpha=0.2), alpha("#252525", alpha=0.2)))

unranked_receptor_fill_colors

A vector of the low and high colors to use for the unranked receptors when show_all_datapoints is TRUE (default: c(alpha("#FFFFFF", alpha=0.2), alpha("#252525", alpha=0.2)))

...

Additional arguments passed to ggplot2::theme. As there are often issues with the scales legend, it is recommended to change legend sizes and positions using this argument, i.e., legend.key.height, legend.key.width, legend.title, and legend.text.

Value

A ggplot object

Details

If the values range of the column used as the "size" parameter is not between 0 and 1.001, an error will be thrown.

The sender cell types can be ordered by encoding the "sender" column as a factor. If the "sender" column is not a factor, the sender cell types will be ordered alphabetically.

By default, the top_n ligand-receptor pairs are shown despite their absolute ranking. So, if a receiver cell type has LR pairs that are only ranked from 31-40 and the top_n is set to 20, the LR pairs will be shown. If use_absolute_rank is set to TRUE, only LR pairs with absolute ranking from 1-20 will be shown.

Examples

if (FALSE) { # \dontrun{
# Create a prioritization table
prior_table <- generate_prioritization_tables(processed_expr_table, processed_DE_table, ligand_activities, processed_condition_markers, prioritizing_weights)
make_mushroom_plot(prior_table)


# Show only top 20, and write rankings on the plot
make_mushroom_plot(prior_table, top_n = 20, show_rankings = TRUE)

# Show all datapoints, and use true color range
make_mushroom_plot(prior_table, show_all_datapoints = TRUE, true_color_range = TRUE)


# Change the size and color columns
make_mushroom_plot(prior_table, size = "pct_expressed", color = "scaled_avg_exprs")


# For a prioritization table with multiple receiver cell types
make_mushroom_plot(prior_table_combined %>% filter(receiver == celltype_oi))
} # }