Chi-square test residual plot — ResidualPlot • Kandinsky

Create a heatmap reporting standardized residuals from chi-square test between two variables of interest

ResidualPlot(
  seurat = NULL,
  var.1 = NULL,
  var.2 = NULL,
  limits = c(-20, 20),
  pval = T,
  perm = 1000,
  padj.thresh = 0.01,
  alternative = c("two.sided", "less", "greater"),
  return.df = F
)

Arguments

seurat: a Seurat object containing Kandinsky data
var.1: character string, name of the first variable
var.2: character string, name of the second variable
limits: lower and upper residual caps for heatmap colour gradient
pval: boolean, whether returning heatmap with cells annotated for significance or not (based on normal distribution fitting)
perm: numeric, number of var.1 and var.2 reshuffling to estimate enrichment pvalue between var.1 and var.2 via permutation test. Default is set to 1000
padj.thresh: numeric, adjusted pvalue threshold to define significant var.1/var.2 chisquare residuals. Default value is set to 0.01
alternative: character string specifying the alternative hypothesis, must be one of "two.sided", "greater" (default) or "less".
return.df: boolean, whether returning only residual plot (FALSE) or also a dataframe containing permutation-based chi-squared test results

Value

residual heatmap created with ggplot2