UpSet or Venn plot of shared taxonomic values across phyloseq objects

Create an UpSet plot (or Venn diagram) showing the shared taxonomic values at a specified rank across all phyloseq objects in a list_phyloseq.

Usage

upset_lpq(x, tax_rank = "Genus", plot_type = "auto", remove_na = TRUE, ...)

Arguments

x: (required) A list_phyloseq object.
tax_rank: (character, required) The name of the taxonomic rank column present in the @tax_table slot of each phyloseq object. For example, "Genus", "Family", or "Species".
plot_type: (character, default "auto") Type of plot to generate. One of "auto", "upset", or "venn". If "auto", uses Venn diagram for 4 or fewer phyloseq objects, UpSet plot otherwise.
remove_na: (logical, default TRUE) If TRUE, remove NA values from the taxonomic rank before computing intersections.
...: Additional arguments passed to ComplexUpset::upset() or ggVennDiagram::ggVennDiagram().

Value

A ggplot2 object (both UpSet and Venn diagrams)

Details

This function extracts the unique values for the specified taxonomic rank from each phyloseq object and creates a visualization showing the intersections between them. UpSet plots are generally better for visualizing complex intersections with more than 4 sets, while Venn diagrams work well for 2-4 sets.

Examples

data("enterotype", package = "phyloseq")
lpq <- list_phyloseq(list(
  fung = data_fungi,
  fung_mini = data_fungi_mini,
  fung_rarefy = rarefy_even_depth(data_fungi),
  enterotype = enterotype
))
#> You set `rngseed` to FALSE. Make sure you've set & recorded
#>  the random seed of your session for reproducibility.
#> See `?set.seed`
#> ...
#> 1016OTUs were removed because they are no longer 
#> present in any sample after random subsampling
#> ...
#> ℹ Building summary table for 4 phyloseq objects...
#> ℹ Computing comparison characteristics...
#> ℹ Checking sample and taxa overlap...
#> ℹ Detected comparison type: NESTED_ROBUSTNESS
#> ℹ 0 common samples, 0 common taxa
#> ℹ Skipping refseq comparison (not all objects have refseq)
#> ✔ list_phyloseq created (NESTED_ROBUSTNESS)
upset_lpq(lpq, plot_type = "upset")
#> Warning: Using `size` aesthetic for lines was deprecated in ggplot2 3.4.0.
#> ℹ Please use `linewidth` instead.
#> ℹ The deprecated feature was likely used in the ComplexUpset package.
#>   Please report the issue at
#>   <https://github.com/krassowski/complex-upset/issues>.

lpq2 <- list_phyloseq(list(
  fung = data_fungi,
  fung_mini = data_fungi_mini
))
#> ℹ Building summary table for 2 phyloseq objects...
#> ℹ Computing comparison characteristics...
#> ℹ Checking sample and taxa overlap...
#> ℹ Detected comparison type: NESTED_ROBUSTNESS
#> ℹ 137 common samples, 45 common taxa
#> ✔ list_phyloseq created (NESTED_ROBUSTNESS)
upset_lpq(lpq2, tax_rank = "Family")