Visualizing Phylogenetic Trees

ggalign provides direct support for phylo objects from the ape package, allowing you to visualize phylogenetic trees alongside other data in an integrated way.

library(ggalign)
#> Loading required package: ggplot2
#> 
#> Attaching package: 'ggalign'
#> The following object is masked from 'package:ggplot2':
#> 
#>     element_polygon

n <- 100
p <- 3
set.seed(3)
tr <- ape::rcoal(n)
x <- matrix(runif(p * n), n, p)

Note that you can also use stack_discrete() to display the tree in a stacked arrangement rather than in a radial layout. For this documentation, we illustrate examples using circle_discrete() as it is the most commonly used approach.

Basic phylogenetic tree alignment

circle_discrete(x,
    radial = coord_radial(inner.radius = 0.2, rotate.angle = TRUE)
) +
    align_phylo(tr) +
    ggalign() +
    geom_tile(aes(y = .column_index, fill = value)) +
    scale_fill_viridis_c()

The underlying data passed to align_phylo() includes both tip and internal node information. Edge data is stored as a ggalign attribute, and you can retrieve it using ggalign_attr(). See ?fortify_data_frame.phylo for details.

circle_discrete(x,
    radial = coord_radial(inner.radius = 0.2, rotate.angle = TRUE)
) +
    align_phylo(tr) +
    geom_point() +
    ggalign() +
    geom_tile(aes(y = .column_index, fill = value)) +
    scale_fill_viridis_c() &
    scale_y_continuous(breaks = NULL)

Grouping by clades

groups <- cutree(as.hclust(tr), k = 3)

By default, align_phylo() will add a single phylogenetic tree, and align_group() can then facet the data by clade:

circle_discrete(x,
    radial = coord_radial(inner.radius = 0.2, rotate.angle = TRUE)
) +
    align_phylo(tr, mapping = aes(color = clade)) +
    geom_point() +
    align_group(groups) +
    ggalign() +
    geom_tile(aes(y = .column_index, fill = value)) +
    scale_fill_viridis_c() &
    scale_y_continuous(breaks = NULL)

Splitting into multiple subtrees

If groups are already set before adding the phylogenetic tree, you can set split = TRUE in align_phylo() to automatically split the tree into multiple subtrees, one for each group.

circle_discrete(x,
    radial = coord_radial(inner.radius = 0.2, rotate.angle = TRUE)
) +
    align_group(groups) +
    align_phylo(tr, mapping = aes(color = clade), split = TRUE) +
    geom_point() +
    ggalign() +
    geom_tile(aes(y = .column_index, fill = value)) +
    scale_fill_viridis_c() &
    scale_y_continuous(breaks = NULL)