ggplot2无法显示次要网格线问题求助

问题说明 #

使用ggplot2绘制瀑布图时，尝试通过panel.grid.minor设置白色次要网格线区分每个tile，但始终无法生效。已尝试单独设置x、y轴网格线，均无效果。

原因分析 #

你的x轴（sampleID）和y轴（gene_name）都是离散型因子变量，ggplot2默认不会为离散轴生成次要网格线，因此panel.grid.minor的设置无法起作用，需要手动生成网格线来实现区分效果。

解决方案 #

通过geom_hline和geom_vline手动添加白色网格线，且需将网格线图层放在geom_tile之前，确保tile显示在网格线上方。

修改后的完整代码 #

library(ggplot2)
library(tidyverse)

# 读取数据（示例用内置数据，实际可替换为read.table("pd_output.txt", header = TRUE, sep = "\t")）
data <- tibble(
  sampleID = c("P-0028", "P-0029", "P-0030", "P-0031", "P-0032", "P-0033", "P-0034", "P-0035", "P-0036"),
  gene_name = c("NCOR1", "SETD2", "ATM", "CDKN1B", "KMT2C", "FOXA1", "NCOR1", "KMT2A", "KMT2C"),
  mutation_types = factor(c("CLONAL", "CLONAL", "SUBCLONAL", "CLONAL", "CLONAL", "CLONAL", "CLONAL", "CLONAL", "CLONAL")),
  variant_consequences = factor(c("missense_variant", "splice_donor_variant", "stop_gained", "frameshift_variant", "stop_gained", "stop_gained", "missense_variant", "missense_variant", "missense_variant")),
  impact = factor(c("MODERATE", "HIGH", "HIGH", "HIGH", "HIGH", "HIGH", "MODERATE", "MODERATE", "MODERATE")),
  clinical_annotations = factor(c("localised", "localised", "localised", "localised", "metastatic", "metastatic", "metastatic", "localised", "localised")),
  TE_fusion = factor(c("no", "yes", "no", "yes", "no", "yes", "yes", "yes", "yes")),
  hotspot = factor(c("no", "yes", "no", "no", "no", "yes", "no", "no", "no"))
)

# 创建瀑布图函数
create_waterfall_plot <- function(data) {
  # 获取离散轴的索引位置，用于生成网格线
  x_positions <- seq_along(unique(data$sampleID))
  y_positions <- seq_along(unique(data$gene_name))
  
  plot <- ggplot(data, aes(x = sampleID, y = gene_name)) +
    theme_bw() +
    # 添加白色分隔网格线（刚好落在tile之间）
    geom_vline(xintercept = x_positions + 0.5, color = "white", size = 2) +
    geom_hline(yintercept = y_positions + 0.5, color = "white", size = 2) +
    theme(
      panel.grid.major = element_blank(),
      panel.grid.minor = element_blank(), # 关闭默认次要网格避免冲突
      axis.text.x = element_text(angle = 90, hjust = 1, vjust = 0.5)
    ) +
    geom_tile(aes(fill = variant_consequences, colour = mutation_types, alpha = 0.5), size = 0.5, width = 0.8, height = 0.8) +
    geom_point(aes(shape = mutation_types, colour = impact), size = 3) +
    scale_fill_manual(values = c(
      "missense_variant" = "blue", 
      "splice_donor_variant" = "orange", 
      "stop_gained" = "darkgreen", 
      "frameshift_variant" = "yellow", 
      "inframe_deletion" = "brown", 
      "missense_variant&splice_region_variant" = "violet", 
      "stop_gained & inframe_deletion" = "gray", 
      "inframe_insertion" = "cyan"
    )) +
    scale_color_manual(values = c("MODERATE" = "lightpink", "HIGH" = "red")) +
    labs(
      x = "Sample ID", 
      y = "Gene Name",
      fill = "Variant Consequences", 
      colour = "Impact", 
      shape = "CLONALITY"
    ) +
    guides(alpha = FALSE) 

  return(plot)
}

# 生成并显示瀑布图
waterfall_plot <- create_waterfall_plot(data)
print(waterfall_plot)

关键修改点 #

1. 网格线位置计算：用seq_along()获取离散轴的索引，通过索引+0.5确定线条位置，刚好落在相邻tile的分界处。
2. 图层顺序：将网格线图层放在geom_tile之前，确保网格线不会被tile遮挡。
3. 关闭默认网格：设置panel.grid.minor = element_blank()，避免默认网格与手动添加的线条冲突。

内容的提问来源于stack exchange，提问作者Venkatesh Chellappa