This tutorial is to visualize microbiome biomarkers with different plots.
In this tutorial, we will use function ggdotchart from R package ggpubr to visualize LefSe biomarkers associated with the number of partners in MSM individuals.
Open a new working R script, and load our example biomarkers from path_to_the_package/KunDH-2023-CRM-MSM_metagenomics/examples/.
>npartner_lefse_df <- data.frame(read.csv("path_to_the_package/KunDH-2023-CRM-MSM_metagenomics/examples/npartners_lefse_deviation_plot.tsv",
header = TRUE,
sep = "\t"))
Use ggdotchart implemented in ggpubr for visualization.
>library(ggpubr)
>ggdotchart(npartner_lefse_df, x = "feature", y = "lda_score",
color = "class",
palette = c("#0073C2FF", "#0073C2FF")
sorting = "descending",
add = "segments",
add.params = list(color = "lightgray", size = 1.5),
group = "class",
rotate = T,
dot.size = 4,
shape = "class",
ggtheme = theme_pubr()
) + theme(text = element_text(size = 13, family = "Arial")) + scale_x_discrete(position = "top")
In this section, we will show you how to visualize LefSe biomarkers associated with multiple groups using UpSet plot.
Our example data comprises LefSe biomarkers associated with sexual practices including RAI: Yes (receiving anal intercourse), having >3 sexual partners (# partners: >3), practicing oral sex (Oral sex: Yes), diagnosed with sexually transmitted infection (STI: Positive), condomless during RAI (Condom use (during RAI): No).
First of all, open a new R working script, and load our example data from path_to_the_package/KunDH-2023-CRM-MSM_metagenomics/examples/.
>library("ComplexHeatmap")
>upset_matrix <- data.frame(read.csv("path_to_the_package/KunDH-2023-CRM-MSM_metagenomics/example_data/UpSet_matrix1.tsv",
header = TRUE,
sep = "\t"))
>rownames(upset_matrix) <- upset_matrix[, colnames(upset_matrix)[[1]]]
>upset_matrix[, colnames(upset_matrix)[[1]]] <- NULL
>upset_matrix <- upset_matrix[as.logical(rowSums(upset_matrix != 0)),] # This step is optional.
Once the data is loaded, we use function UpSet() implemented in the package ComplexHeatmap to draw an UpSet plot.
comb <- make_comb_mat(upset_matrix, mode = "intersect") # generate combination data
c_size <- comb_size(comb) # find combination sizes for setting order later
sets = c("RAI.yes", "X.partners.3", "Oral.yes", "STI.positive", "condom.no") # manually set the set order
upset_plot <- ComplexHeatmap::UpSet(comb,
comb_col = "#fb5238", # the color for combination columns
bg_col = "#ffbeab", # the color for background of columns
bg_pt_col = "#ffdfd5", # the color for background of column patches
set_order = sets,
comb_order = order(c_size),
top_annotation = HeatmapAnnotation(
"# shared taxonomic biomarkers" = anno_barplot(c_size,
ylim = c(0, max(c_size)*1.1),
border = FALSE,
gp = gpar(fill = "#fb5238", col = "#fb5238"),
height = unit(8, "cm")),
annotation_name_side = "left",
annotation_name_rot = 90),
right_annotation = NULL
)
Once the backbone is generated, next we will display the values on the barplot using function decorate_annotation:
upset_plot <- draw(upset_plot)
col_order <- column_order(upset_plot)
decorate_annotation("# shared taxonomic biomarkers", {
grid.text(c_size[col_order], x = seq_along(c_size), y = unit(c_size[col_order], "native") + unit(2, "pt"),
default.units = "native", just = c("left", "bottom"),
gp = gpar(fontsize = 6, col = "#404040"), rot = 45)
})
There are way more features to explore around for making an UpSet plot, please visit the manual.
Similar codes can be used for another set of LefSe biomarkers associated with: RAI: No (not receiving anal intercourse), having 0-3 sexual partners (# partners: 0-3), not practicing oral sex (Oral sex: No), free from sexually transmitted infection (STI: Negative), use condom during RAI (Condom use (during RAI): Yes).
We can generate a nice plot showing the mutual biomarkers shared between different sexual practices by combining these plots.
In the last section, we will use lillipop plot to show the number of shared biomarkers associated with different sexual practices. Here, you can start directly from our prepared file shared_biomarkers.tsv containing the number of LefSe biomarkers shared by different sexual practices (categorized as risk.increasing and risk.reducing).
Firstly, open a new R working script, and load shared_biomarkers.tsv from path_to_the_package/KunDH-2023-CRM-MSM_metagenomics/examples/.
>library(ggpubr)
>shared_biomarkers <- data.frame(read.csv("path_to_the_package/KunDH-2023-CRM-MSM_metagenomics/example_data/shared_biomarkers.tsv",
header = TRUE,
sep = "\t"))
Once the data was loaded, we use a function ggdotchart implement in ggpubr for visualization.
>ggdotchart(shared_biomarkers, x = "group.number", y ="shared.biomarker.number",
color = "type", palette = c("#fb5238", "#469537"), size = 5,
add = "segment",
shape = 19,
group = "cate",
add.params = list(color = "lightgray", size = 2.5),
position = position_dodge(0.25),
ggtheme = theme_pubclean()
) + geom_text(
aes(label = n_common_sps, group = cate),
position = position_dodge(0.8),
vjust = -0.5, size = 3.5
)
Note: The figures displayed above had been edited and arranged using inkscape on the base of the crude output in order to enhance the readability and aesthetic sense.