The TCR_circos function is meant to create a circos-style ggplot full of annotations for single cell TCR data. Each "spoke" is intended to be a clonotype with the outmost ring being a clonotype's frequency and the inner rings showing the frequency of annotations for each clonotype. TCR_circos relies on cowplots to stack the annotation rings.
It takes the following arguments:
df: the data frame off annotations. It is intended to run from a dataframe where all column annotations are factors and each row is a cell.
clonotype: is the (string) name of the column for the outmost clonotype annotations.
annotation: vector of the names of the columns to use as the inner annotation rings.
colors: is a named list of named vectors for the annotation colors.
scale: scale value for the width of the inner rings compared to the outmost. Sensible values are from 0.1 to 0.5.
gap: size of the gap between the rings. Sensible values are from 0 to 0.2.
max.n: The value to scale the count frequency for each clonotype to. Defaults to the max in the data, but can be set to another value so as to keep two separate plots on the same scale.
The function outputs a list of two plots. The first is the circos plot and the second is its legends.
df = read_csv("path/to/example_TCR_data.csv", col_types = "ffffff")
colors = list("Cluster" = rainbow(11)[c(1,6,2,7,8,3,11,4,9,5,10)] %>% setNames(1:11),
"mait_evidence" = topo.colors(12)[c(1,9,5,11,3,7)] %>% setNames(levels(df$mait_evidence)),
"n_TRB" = c("0" = "black", "1" = "tomato", "2"="goldenrod2"))
example_plot1 = TCR_circos(df[1:500,], clonotype ="cdr3s_nt", annotations = annotations, colors = colors)
plot_grid(example_plot1$plot, example_plot1$legend, nrow=1)
example_plot2 = TCR_circos(df[1:500,], clonotype ="cdr3s_nt", annotations = c("mait_evidence","n_TRB"), scale = 0.5, gap = 0.2, colors = colors)
plot_grid(example_plot2$plot, example_plot2$legend, nrow=1) -> example_plot2_with_legend
Currently the plot only orders by clonotype frequency, but later versions will hopfully allow for custom ordering and a dendrogram plot in the middle.
The example data file provided here was downloaded from 10X genomics and processed as if it was the cell metadata for single cell TCR/transcriptome analysis.