volcano_plot: Make a volcano plot comparing the gene expression of selected...

In Close-your-eyes/scexpr: Lots of helpers and wrappers to analyze scRNAseq data sets

Make a volcano plot comparing the gene expression of selected groups of cells from one or more preprocessed Seurat objects

Description

Make a volcano plot comparing the gene expression of selected groups of cells from one or more preprocessed Seurat objects

Usage

volcano_plot(
  SO,
  assay = "RNA",
  volcano.data = NULL,
  negative.group.cells,
  positive.group.cells,
  negative.group.name = "negative.group",
  positive.group.name = "positive.group",
  meta.cols = NULL,
  x.axis.symmetric = T,
  x.axis.extension = 0,
  y.axis.pseudo.log = F,
  pseudo.log.sigma = 1,
  inf.fc.shift = 2,
  pt.size = 1,
  pt.alpha = 0.8,
  font.size = 14,
  pval.tick = NULL,
  min.pct = 0.1,
  max.iter = 10000,
  max.overlaps = 50,
  label.neg.pos.sep = T,
  label.col = "black",
  label.face = "italic",
  font.family = "sans",
  label.size = 4,
  labels.topn = 30,
  label.features = NULL,
  topn.metric = c("p.value", "fc", "both"),
  nudge.x = 0,
  nudge.y = 0,
  p.plot = c("adj.p.val", "p.val"),
  p.adjust = c("bonferroni", "holm", "hochberg", "hommel", "BH", "BY", "fdr", "none"),
  p.cut = NA,
  p.signif = 0.001,
  fc.cut = NA,
  features.exclude = NULL,
  save.path.interactive = NULL,
  gsea.param = NULL,
  interactive.only = F,
  use.limma = F,
  ...
)

Arguments

SO	one or more Seurat object(s); provide multiple objects as a (named) list
assay	which assay to get expression data from
volcano.data	optionally provide a data frame with data that have been calculated before with this function. e.g. if only style element of the returned plot are to be modified, providing this data frame will avoid to re-calculate DE genes
negative.group.cells	vector of cell names for the negative group of cells; genes which are expressed at higher level in this group will have a negative sign (minus); use Seurat::Cells() or Seurat::WhichCells() to select or use filter operations on [e-mail] to select cells
positive.group.cells	vector of cell names for the positive group of cells; genes which are expressed at higher level in this group will have a positive sign (plus); use Seurat::Cells() or Seurat::WhichCells() to select or use filter operations on [e-mail] to select cells
negative.group.name	name for the negative group of cells; this will appear on plot axes
positive.group.name	name for the positive group of cells; this will appear on plot axes
meta.cols	which meta.cols to keep in SO for interactive plotting
x.axis.symmetric	logical; make the x-axis which indicates log-fold changes symmetric (same range in plus and minus direction)
x.axis.extension	numeric; positive or negative value to extend the x-axis by
y.axis.pseudo.log	logical; use a pseudo-log y-axis for -log10(pvalue); may be helpful to better visualize and label dense area of genes in the plot; done with scales::pseudo_log_trans
pseudo.log.sigma	numeric; adjustment to the pseudo-log tranasformation; passed to scales::pseudo_log_trans
inf.fc.shift	numeric; genes with an infinite fold-change (happens when a gene not expressed at all in one group) are not left infinite but will be put at: "largest finite fold-change +/- inf.fc.shift"; such genes are indicated specifically
pt.size	numeric; point size passed to ggplot2::geom_point
pt.alpha	numeric; point opacity (alpha value) passed to ggplot2::geom_point
font.size	numeric; font size of point labels; passed as base_size to ggplot2::theme_bw() and as size = 5/14*font.size to ggplot2::geom_text (5/14 is the conversion between how geom_text handles the size argument and how theme does it)
pval.tick	NULL or numeric, if not NULL: plot an extra y-axis-tick to indicate a significance level in the linear space (not -log10); e.g.: pval.tick = 0.01 will plot an axis-tick at p = 0.01
min.pct	numeric; only consider genes for the volcano plot which are expressed at least by this fraction of cells in both groups (negative.group.cells, positive.group.cells)
max.iter	max.iter passed to ggrepel::geom_text_repel
max.overlaps	max.overlaps passed to ggrepel::geom_text_repel
label.neg.pos.sep	logical whether to label genes with negative and positive fold change separately; this will make sure that labels on the left (negative log fc) point to the left and labels on the right (positive log fc) point to the right
label.col	font color of labels
label.face	font face of labels
font.family	font family (font type) of labels, e.g. Courier or Sans
label.size	size of labels
labels.topn	numeric; how many labels (roughly) to plot based on the selected metric
label.features	character vector; select genes which are to be labeled; OR: "significant" which will let p.signif come into usage
topn.metric	which metric to use for selection of top genes for labeling ()
nudge.x	nudge.x passed to ggrepel::geom_text_repel; shift labels into x-direction; for genes with negative fold change -nudge.x will be used; genes with positive fold change use nudge.x
nudge.y	nudge.x passed to ggrepel::geom_text_repel; shift labels into y-direction;
p.plot	which p-value to use for plotting, adjusted p-value or unadjusted p-value
p.adjust	method for p-value adjustment
p.cut	plot a horizontal line at this p-value; in combination with fc.cut all genes above this cut (and above fc.cut) are counted and the number is plotted
p.signif	a significance level from which on genes are labeled; supply label.features = "significant" to make use of this
fc.cut	plot vertical lines at these fold changes (plus and minus); in combination with p.cut all genes above this cut (and above p.cut) are counted and the number is plotted
features.exclude	character vector of features to exclude from plotting; you may supply regular expressions like "^RPL" and/or "^RPS" to exclude all ribosomal genes
save.path.interactive	path on disk where to save data for interactive analysis of the volcano plot; this will initiate a directory with a shiny script and an rds file of with SO and DE genes
gsea.param	list of length 2, each index holding a numeric vector of length 2: fold change (list index 1) and p-value (list index 2) limits for GSEA; first entry of each vector applies for genes with negative fold change, second entry for genes with positive fold change; 2 GSEA will be run separately on genes with negative and positive fold change; if argument is NULL no GSEA is performed; the function called for GSEA is scexpr::fgsea_on_msigdbr which uses all gene sets from msigdb(r) by default and runs fgsea for GSEA; alternative arguments to scexpr::fgsea_on_msigdbr can be supplied in ... (except for gene.ranks)
interactive.only	logical; do calculations only for interactive analysis and only return these values
use.limma	logical; use limma for DE gene detection; intended for subsequent use MetaVolcanoR which relies on values returned from limma
...	arguments to scexpr::feature_plot like col.pal.d = setNames(c("grey90", scexpr::col_pal()c(2,3)), c("other", "name1", "name2")), order.discrete = "^other", plot.title = F, etc

Close-your-eyes/scexpr documentation built on April 21, 2023, 10:27 a.m.