代码目的

• 和弦图 (Chord Diagram) 可以清晰地显示复杂的数据交互，尤其适用于展示多个类别之间的关系和交叉。
  
• 外围环绕条形图是对和弦图的补充，这篇文章显示的是不同管理实践类型下的效率，这样就能与和弦图中的变量一起揭示规律了。

代码

setwd("D:/生信代码复现/顶刊和弦图")
library(tidyverse)
library(circlize)
library(grid)
library(circlize)
library(ggplotify)

# Read data
data1 <- readxl::read_xlsx("41586_2024_7020_MOESM4_ESM.xlsx", sheet = "The inner cycle")
data2 <- readxl::read_xlsx("41586_2024_7020_MOESM4_ESM.xlsx", sheet = "The outer cycle")

# 将最后一列设置为行名
last_col_name <- names(data1)[ncol(data1)]
data1 <- data1 %>%
  column_to_rownames(var = last_col_name)

data1_matrix <- as.matrix(data1)

head(data1_matrix)

# 提取物种和类型并保存为新列
data2 <- data2 %>%
  mutate(Species = sub("^([^-]+).*", "\\1", `Ctype-Management practices`),
         Type = sub("^[^-]+-(.*)$", "\\1", `Ctype-Management practices`)) 

head(data2)

color_region = c("#6BCCEF", "#00D187", "#009462", "#FDC848", "#9C9C9C", "#951255")
color_species = c("#810084", "#0075B6", "#000000")

# 和弦图
circos.clear()
circos.par(start.degree = 90, circle.margin = c(0.6, 0.6, 0.6, 0.6))

chordDiagram(data1_matrix, 
             annotationTrack = "grid",
             transparency = 0.5,
             annotationTrackHeight = 0.07,
             preAllocateTracks = list(track.height = 0.005),
             big.gap = 1, small.gap = 1,
             #scale = T,
             grid.col = c(rep(color_region, 3), rep("#525252", 9)))
# 添加扇区名称并控制方向
circos.track(track.index = 1, panel.fun = function(x, y) {
  # 提取最后的部分
  sector_name <- CELL_META$sector.index
  sector_name <- sub(".*-", "", sector_name)  # 去掉分隔线前的部分
  circos.text(CELL_META$xcenter, CELL_META$ylim[2], sector_name, 
              facing = "clockwise", niceFacing = TRUE, adj = c(0, 0.5), cex = 0.75)
}, bg.border = NA)

###-----------------------------------------------------------------------------
##'@写一个函数，将绘图代码存放在函数中，方便输出图形
create_chord_plot <- function(data1_matrix, data2, color_region) {
  circos.clear()
  circos.par(start.degree = 90, circle.margin = c(0.6, 0.6, 0.6, 0.6))
  
  chordDiagram(data1_matrix, 
               annotationTrack = "grid",
               transparency = 0.5,
               annotationTrackHeight = 0.07,
               preAllocateTracks = list(track.height = 0.005),
               big.gap = 1, small.gap = 1,
               #scale = T,
               grid.col = c(rep(color_region, 3), rep("#525252", 9)))
  # 添加扇区名称并控制方向
  circos.track(track.index = 1, panel.fun = function(x, y) {
    # 提取最后的部分
    sector_name <- CELL_META$sector.index
    sector_name <- sub(".*-", "", sector_name)  # 去掉分隔线前的部分
    circos.text(CELL_META$xcenter, CELL_META$ylim[2], sector_name, 
                facing = "clockwise", niceFacing = TRUE, adj = c(0, 0.5), cex = 0.75)
  }, bg.border = NA)
}

###--------------------------------------------------------------------
# 重新设置par参数
par(mar = c(1, 1, 1, 1))  # 设置边距
par(asp = 1)  # 强制设置图形为正方形

# 创建和弦图并转换为 ggplot 对象
p1 <- as.ggplot(expression({
  create_chord_plot(data1_matrix, data2, color_region)
}))

# 显示图形
print(p1)


p1
# 在每个物种内部按 `EF_mean (%)` 排序 `Ctype-Management practices`
data2 <- data2 %>%
  group_by(Species) %>%
  arrange(Species, desc(`EF_mean (%)`), `Ctype-Management practices`) %>%
  ungroup() %>%
  mutate(ID = seq(1, nrow(data2)))


# 设置添加label标签信息
# Get the name and the y position of each label
angle <- 90 - 360 * (data2$ID-0.5) /nrow(data2)   
data2$hjust <- ifelse( angle < -90, 1, 0)
data2$angle <- ifelse(angle < -90, angle+180, angle)

p2 <- ggplot(data2, aes(x = ID, 
                        y = `EF_mean (%)`)) + 
  geom_bar(stat = "identity", aes(fill = Species)) + 
  coord_polar() +
  ylim(-60, max(data2$`EF_mean (%)`)+6) + 
  scale_fill_manual(values = color_species) +
  theme_void() + 
  theme(legend.position = "none",
        plot.margin = unit(c(-1,-1,-1,-1), "cm")) +
  geom_text(aes(y = `EF_mean (%)`+1, label = Type, hjust=hjust), 
            angle= data2$angle, size = 3.2) +
  geom_text(aes(y = `EF_mean (%)`/2, label = sprintf("%.1f", `EF_mean (%)`)), 
            angle= data2$angle, size = 3.5, color = "white") 
p2


# Add legends
legend(x = -1.855, y = 1.8, , legend = c("Maize", "Rice","Wheat"), fill = color_species, 
       border = "white", bty = "n", x.intersp = 0.2, y.intersp = 0.55,
       title = "Outer circle", title.adj = 0.4, title.font = 2,
       cex = 0.75)
legend(x = -1.82, y = 1.5, legend = c("South America", 
                                      "Africa", "Asia", 
                                      "Oceania", "North America", "Europe", "Driver"), 
       fill = c(color_region, "#525252"), 
       border = "white", bty = "n", x.intersp = 0.2, y.intersp = 0.55,
       title = "Inner circle", title.adj = 0.16, title.font = 2,
       cex = 0.75)


library(cowplot)
library(grid)

ggdraw()+
  draw_plot(p2)+
  draw_plot(p1,x = 0, y = 0,
            width = 1, height = 1)

ggsave("./20241127_和弦图.jpg",width = 10,height = 10)

复现出图