Adding EnhancedVolcano function to the package

[lobanovav]

A quick description: Enhanced Volcano is now a function into the R folder and created a unit test with synthetic data

[Copilot]

Pull request overview

This PR introduces a new Volcano_Plot() function to generate static + interactive volcano plots (via EnhancedVolcano + plotly) and adds a testthat unit test suite using synthetic DEG data.

Changes:

• Added R/Volcano_Plot.R implementing volcano plot generation for one or multiple comparisons, returning plots plus ranked output data.
• Added tests/testthat/test-Volcano_Plot.R covering basic execution, return structure, multiple comparisons, and a few edge cases.

Reviewed changes

Copilot reviewed 2 out of 2 changed files in this pull request and generated 8 comments.

File	Description
R/Volcano_Plot.R	New exported plotting function using EnhancedVolcano with optional labeling and interactive plotly output.
tests/testthat/test-Volcano_Plot.R	New tests validating Volcano_Plot() output structure and a few edge cases.

---

💡 Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.

[Copilot]

In the use_custom_lab branch, lfc_name/sig_name can be undefined when change_lfc_name/change_sig_name are non-empty (the typical case), which will error at colnames(df) <- ... and later lookups. Set lfc_name/sig_name deterministically (e.g., to the chosen label names) and ensure they always correspond to actual column names in df.

Suggested change

	if (nchar(change_lfc_name) == 0){lfc_name = lfc.col[i]}
	if (nchar(change_sig_name) == 0){sig_name = sig.col[i]}
	colnames(df) <- c(label.col, change_lfc_name, sig_name)
	} else {
	lfc_name = lfc.col[i]
	sig_name = sig.col[i]
	lfc_name <- if (nchar(change_lfc_name) == 0) lfc.col[i] else change_lfc_name
	sig_name <- if (nchar(change_sig_name) == 0) sig.col[i] else change_sig_name
	colnames(df) <- c(label.col, lfc_name, sig_name)
	} else {
	lfc_name <- lfc.col[i]
	sig_name <- sig.col[i]

[Copilot]

The y-limit capping logic overwrites p-values with maxy (a -log10(p) value): df[[sig_name]][!keep] <- maxy. That produces invalid p-values (>1) and will move highly-significant points to the bottom of the plot. If you want to cap the plotted y, either cap using the corresponding p-value (10^-maxy) or (preferably) leave the data unchanged and apply a plotting limit (e.g., via ggplot scales/coord_cartesian).

Suggested change

	df[[sig_name]][!keep] <- maxy
	df[[sig_name]][!keep] <- 10^(-maxy)

[Copilot]

genes_to_label <- df_sub[1:no_genes_to_label, ...] will introduce NA labels (or error) when df_sub has fewer than no_genes_to_label rows, including when the is_red filter yields 0 rows. Consider using head()/seq_len(min(nrow(df_sub), no_genes_to_label)) and dropping NA from genes_to_label before passing to EnhancedVolcano.

Suggested change

	genes_to_label <- as.character(df_sub[1:no_genes_to_label, label.col])
	n_to_label <- min(nrow(df_sub), no_genes_to_label)
	genes_to_label <- as.character(df_sub[seq_len(n_to_label), label.col])
	genes_to_label <- genes_to_label[!is.na(genes_to_label)]

[Copilot]

The loop for (i in 1:length(lfc.col)) behaves incorrectly when lfc.col is length 0 (1:0), and there’s no validation that sig.col and lfc.col have the same length before indexing both. Prefer seq_along(lfc.col) and add an early argument check that length(sig.col) == length(lfc.col) (and that length > 0).

Suggested change

	for(i in 1:length(lfc.col)){
	# Validate lfc.col and sig.col arguments
	if (missing(lfc.col) || missing(sig.col)) {
	stop("Both 'lfc.col' and 'sig.col' must be provided.")
	}
	if (length(lfc.col) == 0L || length(sig.col) == 0L) {
	stop("'lfc.col' and 'sig.col' must have length greater than 0.")
	}
	if (length(lfc.col) != length(sig.col)) {
	stop("'lfc.col' and 'sig.col' must have the same length.")
	}
	for (i in seq_along(lfc.col)) {

[Copilot]

This function imports and calls EnhancedVolcano::EnhancedVolcano, but the package is not listed in DESCRIPTION (so R CMD check will fail once NAMESPACE is regenerated). Add EnhancedVolcano to Imports (or Suggests + guard usage), and ensure NAMESPACE/man/ are regenerated to include the new @export and imports.

[Copilot]

The interactive plot construction adds a new geom_point() layer on top of p_empty, so points are drawn twice (once from EnhancedVolcano, once from the added layer) and pointSize/other aesthetics may diverge between layers. Consider attaching the text aesthetic to the existing point layer (or modifying/replacing that layer) so the interactive plot doesn’t duplicate geometry and avoids the extra ggplot_build() work.

Suggested change

	# Extract the data used for plotting
	plot_data <- ggplot_build(p_empty)$data[[1]]
	pxx <- p_empty +
	xlab("Fold Change") +
	ylab("Significance") +
	theme_minimal() +
	geom_point(aes(
	text = paste("Gene:", df[[label.col]],
	"<br>Log2FC:", df[[lfc_name]],
	"<br>P-value:", df[[sig_name]]),
	colour = as.character(plot_data$colour),
	fill = as.character(plot_data$colour)
	),
	shape = 21,
	size = 2,
	stroke = 0.1) +
	scale_fill_identity()
	# Add hover text aesthetic directly to the existing plot
	pxx <- p_empty +
	xlab("Fold Change") +
	ylab("Significance") +
	theme_minimal() +
	aes(
	text = paste(
	"Gene:", df[[label.col]],
	"<br>Log2FC:", df[[lfc_name]],
	"<br>P-value:", df[[sig_name]]
	)
	)

[Copilot]

library(dplyr) is not used in this test file. Dropping unused test dependencies helps keep the test environment minimal and avoids masking functions during tests.

Suggested change

library(dplyr)