Plots of MitoCarta 3.0 FGSEA Results

This article generates a scatterplot and heatmap of the MitoCarta 3.0 FGSEA results (Fig. 5A–B).

library(MotrpacRatTraining6moWATData)
library(MotrpacRatTraining6moWAT)
library(dplyr)
library(ggplot2)
library(ComplexHeatmap)
library(latex2exp)

save_plots <- dir.exists(paths = file.path("..", "..", "plots"))

## SED MvF plot ---------------------------------------------------------------
x <- PROT_MITOCARTA_FGSEA$MvF_SED %>%
  filter(padj < 0.05) %>%
  arrange(NES) %>%
  mutate(pathway = factor(pathway, levels = unique(pathway)),
         leading_edge_size = lengths(leadingEdge),
         top_genes = unlist(lapply(leadingEdge_genes, function(x)
           paste(head(x, 6), collapse = ", ")
         )),
         top_genes = ifelse(leading_edge_size > 6,
                            paste0(top_genes, ", ..."), top_genes))

p1 <- ggplot(x, aes(x = NES, y = as.numeric(pathway),
                    size = leading_edge_size)) +
  geom_point(color = "#5555ff", alpha = 0.7, shape = 16) +
  scale_size_area(name = "Leading Edge Size:",
                  max_size = 4) +
  scale_x_continuous(breaks = seq(1.6, 2.2, 0.2),
                     limits = c(1.5, 2.2)) +
  scale_y_continuous(breaks = 1:length(x$pathway),
                     labels = levels(x$pathway),
                     expand = expansion(add = 0.5)) +
  labs(x = latex2exp::TeX("NES: (Male - Female)$_{SED}$"),
       y = NULL) +
  theme_pub() +
  theme(panel.grid.minor = element_blank(),
        # panel.grid.major.x = element_blank(),
        axis.ticks.y = element_blank(),
        axis.text.y.right = element_text(size = 5, color = "black"),
        legend.margin = margin(t = -2, unit = "pt"),
        panel.border = element_rect(color = "black", linewidth = 0.3,
                                    fill = NA),
        axis.line = element_blank(),
        legend.direction = "horizontal", 
        legend.position = "bottom",
        legend.key.size = unit(10, "pt"))

p1

ggsave(file.path("..", "..", "plots", "MitoCarta_FGSEA_MvF.pdf"), p1, family = "ArialMT",
       height = 2.8, width = 2.2, dpi = 400, bg = "white")

## Timewise 4W and 8W comparison heatmap -------------------------------------
scale <- 2
heatmap_legend_side <- annotation_legend_side <- "right"
align_heatmap_legend <- align_annotation_legend <- "heatmap_top"
direction <- "vertical"
legend_width <- NULL
legend_height <- unit(90, "points")

column_labels <- rep(paste0(2 ^ (0:3), "W - SED"), times = 2)

anno_df <- data.frame(Sex = rep(c("Female", "Male"), each = 4))
anno_colors <- list(Sex = c("Female" = "#ff6eff", "Male" = "#5555ff"))

top_annotation <- HeatmapAnnotation(
  df = anno_df,
  col = anno_colors,
  gap = unit(0, "points"),
  gp = gpar(col = "black"),
  annotation_name_gp = gpar(fontsize = 7 * scale),
  annotation_legend_param = list(
    nrow = 2, by_row = T,
    grid_height = unit(14, "points"),
    grid_width = unit(14, "points"),
    border = "black",
    title_gp = gpar(fontsize = 7 * scale),
    labels_gp = gpar(fontsize = 6 * scale)
  )
)
column_split <- anno_df$Sex

colors <- c("#3366ff", "darkred", "grey")

# Get terms that are enriched in later timepoints
late_terms <- PROT_MITOCARTA_FGSEA$trained_vs_SED %>%
  filter(grepl("4|8", contrast)) %>%
  group_by(pathway) %>%
  summarise(any_sig = any(padj < 0.05),
            max_val = max(-log10(padj), na.rm = TRUE)) %>%
  filter(any_sig) %>%
  slice_max(order_by = max_val, n = 20) %>%
  pull(pathway)

PROT_MITOCARTA_FGSEA$trained_vs_SED %>%
  filter(pathway %in% late_terms) %>%
  enrichmat(n_top = 20,
            colors = colors,
            scale_by = "row",
            cell_size = unit(6 * scale , "points"),
            padj_legend_title = "BH-adjusted\np-value",
            padj_args = list(
              title_gp = gpar(fontsize = 7 * scale),
              labels_gp = gpar(fontsize = 6 * scale),
              grid_height = unit(6 * scale, "pt"),
              grid_width = unit(6 * scale, "pt")
            ),
            rownames_column = "pathway",
            heatmap_args = list(
              column_labels = column_labels,
              column_title_side = "bottom",
              column_split = column_split,
              column_title = NULL,
              column_names_gp = gpar(fontsize = 5.5 * scale),
              row_names_gp = gpar(fontsize = 5.5 * scale),
              top_annotation = top_annotation,
              heatmap_legend_param = list(
                title_gp = gpar(fontsize = scale * 6.5),
                labels_gp = gpar(fontsize = scale * 6),
                direction = direction,
                legend_width = legend_width,
                legend_height = legend_height
              )
            ),
            filename = file.path("..", "..", "plots",
                                 "MitoCarta_FGSEA_timewise.pdf"),
            height = scale * unit(2.5, "in"),
            width = scale * unit(3.2, "in"),
            units = "in",
            draw_args = list(
              legend_gap = unit(0.4, "in"),
              gap = unit(0.5, "in"),
              adjust_annotation_extension = TRUE,
              heatmap_legend_side = heatmap_legend_side,
              annotation_legend_side = annotation_legend_side,
              align_annotation_legend = align_annotation_legend,
              align_heatmap_legend = align_heatmap_legend
            ),
            save_args = list(res = 400, dpi = 400, family = "ArialMT")
  )

Tyler Sagendorf

01 May, 2024