NEWS

---

ggvolc 0.4.0

DE objects in, colorblind-safe palettes

• Pass DE result objects directly. ggvolc() and genes_table() now accept the objects themselves, not only data frames:
  • DESeq2: DESeqResults (from results() / lfcShrink()) and a DESeqDataSet after DESeq() (results() is called for you).
  • edgeR: TopTags, and DGELRT / DGEExact from glmQLFTest(), glmLRT() and exactTest() (all genes are kept).
  • limma: MArrayLM fits after eBayes(); choose the coefficient with the new coef argument when the fit has several.
• New exported de_table() does the conversion, for when you want the data frame yourself.
• palette argument with built-in presets, listed by the new ggvolc_palettes(): colorblind-safe "okabe_ito" and "rdbu", and presets from the ltc palettes ("pantone23", "seafarer", "midnight", "maya", "trio1"). palette also takes a named vector such as c(down = "navy", up = "firebrick"). The default colors are unchanged.
• up_reg_color, down_reg_color and not_sig_color now default to NULL and, when given, override the matching palette color.
• genes_table() colors the log2FC column to match the plot's palette, and gains palette and coef arguments. The table now spans the full width of the plot panel with larger text, and takes only the height its rows need (table_height now defaults to NULL; set it to fix the ratio as before).
• top_n in genes_table() selects the most significant genes automatically.
• Compare contrasts. Pass a named list of results, e.g. ggvolc(list("Day 1" = res1, "Day 7" = res7)), to get one panel per contrast on shared axes. label_top and show_counts work per panel, and data2 genes are highlighted in every panel. Panels are titled with clean bold labels (restyle with + theme(strip.background = ...)).
• highlight colors named gene sets (pathways, GO terms) with their own legend entries, in colors matching the default palette; choose colors with highlight_colors and per-set transparency with highlight_alpha.
• show_counts = TRUE prints the number of down- and upregulated genes in the top corners.
• flip = TRUE swaps the axes: log2 fold change on the y-axis, significance on the x-axis. When comparing contrasts side by side this lines the fold-change axis up across panels.
• facet_scales controls which axes compared panels share. By default only the significance axis is free ("free_y", or "free_x" with flip = TRUE), so the fold-change axis stays comparable while a weak contrast is not squashed by a strong one; "fixed" shares both.
• New dataset stress_timecourse: real DESeq2 results (30, 60 and 180 min vs 0) from a fission yeast osmotic stress time course (Leong et al. 2014, via the Bioconductor package fission), for trying out contrast comparison.
• Genes with an NA p-value or fold change (e.g. those DESeq2 removes by independent filtering) are left out with a message, instead of triggering ggplot2's "Removed rows" warning.
• data2 can now be a plain character vector of gene names.
• Text colors in the genes_table() table and the counts are darkened when needed to reach a WCAG contrast ratio of 4.5:1 on white, so light palette colors (e.g. pantone23's gold) stay readable.
• DESeq2, edgeR and limma are optional (Suggests) dependencies.

ggvolc 0.3.0

Better significance handling, auto-labels & interactivity

• Significance on the adjusted p-value (FDR) by default. ggvolc() gains a sig_col argument ("padj" or "pvalue") and now defaults to "padj", so hits are called on the FDR — the recommended cutoff in most DE workflows — with the y-axis and the significance segment following the choice so the plot stays internally consistent. Set sig_col = "pvalue" for the raw p-value. When the default "padj" is used but the data has no adjusted-p column, ggvolc falls back to the raw p-value automatically.
• Robust to p == 0. DESeq2 and edgeR can report a p-value of exactly 0 (floating-point underflow) for the strongest genes; -log10(0) is Inf, which ggplot2 silently drops — so the most significant genes used to vanish from the plot. Those values are now capped to a finite ceiling (just above the most significant real gene) and a message reports how many were adjusted.
• label_top = N. Automatically highlight and label the N most significant genes without building a separate data2. Combines with an explicit data2 when both are supplied.
• label_dir. Choose the direction the labelled genes are drawn from: "both" (default), "up", "down", or "each" (top N upregulated and top N downregulated). Pairs with label_top, e.g. label_top = 10, label_dir = "each".
• title argument. The plot title is now configurable and defaults to NULL (no title), replacing the previous hardcoded placeholder title.
• Interactive volcano plots. ggvolc(..., interactive = TRUE) returns a ggiraph girafe widget where hovering a point reveals the gene name and its statistics. ggiraph is an optional (Suggests) dependency, so the core package still installs without it.
• New internal helper: neglog10_cap().
• The README is now generated from README.Rmd (via devtools::build_readme()) so every example plot is rendered from live code and kept in sync with the package.
• DESCRIPTION now declares R (>= 4.1.0), matching the native pipe (|>) used internally (previously mis-declared as >= 3.5.0).

ggvolc 0.2.0

Multi-pipeline support & gt tables

• DESeq2, edgeR, and limma auto-detection: ggvolc() and genes_table() now accept output from all three major DE pipelines. Column names are automatically mapped internally — no manual renaming needed.
  • DESeq2: log2FoldChange, pvalue, padj, baseMean
  • edgeR: logFC, PValue, FDR, logCPM
  • limma: logFC, P.Value, adj.P.Val, AveExpr
• Gene identifiers stored as row names (common in edgeR/limma) are automatically promoted to a genes column.
• genes_table() rewritten with gt + patchwork:
  • Replaces gridExtra::grid.arrange with patchwork::wrap_table()
  • Gene table rendered as a gt table with formatted numerics, color-coded p-values, and directional log2FC coloring
  • p-value / adjusted-p columns use grid-safe scientific notation (1.0e-08) so exponents render correctly once the table is composed into the plot graphic by patchwork
  • Returns a proper patchwork object (composable, ggsave()-able)
• New internal helpers: standardize_de_columns(), detect_de_source()
• Added testthat test suite covering all three DE formats, column detection, attention genes, plot options, and the new gene table.
• Dependencies: replaced gridExtra with gt and patchwork.
• Version bump to 0.2.0.

ggvolc 0.1.0 (2026-01-16)

Initial CRAN Release

• Initial release of ggvolc package
• Create customizable volcano plots for differential gene expression analysis
• Functions included:
  • ggvolc(): Main function to create volcano plots
  • genes_table(): Combine volcano plots with gene tables
• Features:
  • Highlight genes of interest
  • Adjustable significance thresholds (p-value and fold change)
  • Customizable colors for up/down regulated genes
  • Optional significance segment lines
  • Size scaling by log2FoldChange or p-value
  • Integration with ggplot2 for additional customization
• Example datasets included:
  • all_genes: Complete differential expression results
  • attention_genes: Subset of genes to highlight