NEWS

CanvasXpressBio 0.99.4

Fix cxvolcano(): the result matrix was transposed, so every gene collapsed onto two plotted points. Genes are now the variables and the fold-change / p-value are the axis columns, giving one point per gene.
cxvolcano() now colours points by regulation (up / down / not significant), draws the fold-change and p-value threshold lines, and gains label, label_genes, label_fc and label_p to label significant or named genes.
cxplot() gains overlays, var_overlays and cluster to draw colData / rowData annotations as overlay tracks and to cluster rows and columns.
New cxsurvival() draws a Kaplan-Meier curve from colData.
New cxboxplot() draws a paired boxplot of one feature with per-subject connecting lines, faceting and colouring from colData.
Rewrote the vignette around a single SummarizedExperiment shown as a heatmap, volcano, Kaplan-Meier curve and paired boxplot, and added a section motivating the package relative to existing Bioconductor visualization tools.

Version bump to re-run the Bioconductor staging build (no code changes; 0.99.2 built and passed checks but failed to publish).

Vignette charts now render in the built HTML: require canvasXpress (>= 1.70.3), which ships a CanvasXpress library that BiocStyle's self-contained HTML post-processing no longer corrupts (earlier versions were mangled on inlining, leaving the widgets blank). Verified by rendering the vignette against the CRAN 1.70.3 library.

Initial submission to Bioconductor.
cxplot() generic renders a SummarizedExperiment (and extending classes such as SingleCellExperiment and DESeqDataSet) as an interactive CanvasXpress heatmap, with colData as sample annotations and rowData as feature annotations. n= keeps only the most variable features.
cxvolcano() draws a volcano plot from a differential-expression result table (DESeqResults, limma::topTable(), or a data frame).