Plot a Euclidean Distance Heatmap

Compute a sample-by-sample distance matrix from a geneda object and plot a ggplot2 heatmap. Optionally subset metadata columns and reorder rows/columns by hierarchical clustering.

PlotDistances(
  object,
  method = "euclidean",
  reorder = TRUE,
  meta_cols = NULL,
  palettes = NULL,
  return = c("object", "plot")
)

Arguments

object

A geneda object

method

Distance method for stats::dist (default "euclidean")

reorder

Logical; if TRUE, reorder rows/cols by hclust on the distance matrix. Default TRUE

meta_cols

Optional character vector of metadata columns to display in the title/subtitle for quick reference (no annotation strips are drawn).

palettes

Optional named list of color vectors that can be used downstream (returned) if the caller wants to annotate elsewhere.

return

One of c("object", "plot"). If "plot", draws the heatmap immediately via grid and still returns the result list. Default "object".

Value

A list with elements: dist_matrix, order (character vector of sample names), heatmap (pheatmap object), palettes (as passed through)

Examples

# \donttest{
mock_norm <- matrix(rnorm(12, mean = 0, sd = 2), nrow = 4, ncol = 3)
colnames(mock_norm) <- paste0("Sample", 1:3)
rownames(mock_norm) <- paste0("Gene", 1:4)
mock_meta <- data.frame(condition = c("A","B","A"), row.names = colnames(mock_norm))
obj <- GenEDA(normalized = mock_norm, metadata = mock_meta)
colorList <- list(condition = c("A" = "red", "B" = "blue"))
PlotDistances(
    obj,
    meta_cols = c("condition"),
    palettes = colorList,
    return = "plot")

#> $dist_matrix
#>          Sample1  Sample2  Sample3
#> Sample1 0.000000 5.401900 5.935052
#> Sample2 5.401900 0.000000 3.425435
#> Sample3 5.935052 3.425435 0.000000
#> 
#> $order
#> [1] "Sample1" "Sample2" "Sample3"
#> 
#> $heatmap
#> 
#> $palettes
#> $palettes$condition
#>      A      B 
#>  "red" "blue" 
#> 
#> 
    # }