Phage type

Genome molecule type

Bacteriophage genomes can be compose of DNA or RNA. In the NCBI the majority of bacteriophages have a DNA genome (97.1%) followed by ss-DNA genomes (2.7%).

Show the code

library(tidyverse)
library(patchwork)

df <- read_tsv("rawdata/clean_genomes_data.tsv") %>%
    select(molecule, jumbophage) %>%
    count(molecule, jumbophage) %>%
    mutate(
        percent = 100 * (n / sum(n)),
        jumbophage = if_else(jumbophage == FALSE, "Phage",
            "Jumbophage"
        ),
        molecule = factor(molecule,
            levels = c("DNA", "ss-DNA", "RNA", "ds-RNA", "ss-RNA", "cRNA", "mRNA")
        )
    )

base_color <- c("#d8b365", "#5ab4ac")

p1 <- df %>%
    ggplot(aes(
        x = molecule,
        y = percent,
        fill = jumbophage
    )) +
    geom_col() +
    scale_y_continuous(
        expand = c(0, 0),
        limits = c(0, 100),
        breaks = seq(0, 100, 10)
    ) +
    scale_fill_manual(values = base_color) +
    labs(
        y = "Number of phages (%)",
        x = NULL
    ) +
    theme(
        text = element_text(size = 20),
        panel.background = element_blank(),
        panel.border = element_rect(
            colour = "black", fill = NA, linewidth = 1
        ),
        legend.position = "bottom",
        legend.title = element_blank()
    )


p2 <- df %>%
    filter(!molecule == "DNA" & !molecule == "ss-DNA") %>%
    ggplot(aes(
        x = molecule,
        y = percent,
        fill = jumbophage
    )) +
    geom_col() +
    scale_y_continuous(
        expand = c(0, 0),
        limits = c(0, 0.2),
        breaks = seq(0, 0.2, 0.05)
    ) +
    scale_fill_manual(values = "#5ab4ac") +
    labs(
        y = "Number of phages (%)",
        x = NULL
    ) +
    theme(
        text = element_text(size = 12),
        panel.background = element_blank(),
        panel.border = element_rect(
            colour = "black", fill = NA, linewidth = 0.5
        ),
        legend.position = "none"
    )

p1 +
    inset_element(p2, left = 0.35, bottom = 0.4, right = 0.95, top = 0.95)