R语言绘制自定义参数红色系代谢物相关性热图求助

采用ComplexHeatmap包实现所有定制化需求，代码可直接运行，适配大规模数据集。

第一步：安装加载依赖包 #

首次运行需先安装依赖，后续使用直接加载即可：

# 首次使用执行安装
if (!require("BiocManager", quietly = TRUE))
    install.packages("BiocManager")
BiocManager::install("ComplexHeatmap")
install.packages(c("circlize", "tidyverse"))

# 加载包
library(ComplexHeatmap)
library(circlize)
library(tidyverse)

第二步：数据预处理 #

将输入数据转为热图要求的矩阵格式，同时配置Y轴亚部分组信息：

# 加载示例数据
data.str <-
  structure(
    list(
      Metabolite = c(
        "Glucose_%",
        "Fructose_%",
        "Sugars_%",
        "Sugars.as.mono_%",
        "Starch_%",
        "Starch.as.mono_%",
        "Glutamic_%",
        "Proline_%",
        "Biotin_%",
        "C20.2cn6_%"
      ),
      Difference_S_HDL_CE_pct = c(
        27.0580967,
        29.4741588,
        30.6727965,
        37.044534,
        0.00592,
        0.00924,
        2.8089622,
        4.3284966,
        3.616572,
        16.6234106
      ),
      Difference_L_HDL_TG = c(
        20.3426932,
        19.6578323,
        6.1540709,
        6.618907,
        20.7397107,
        22.7363078,
        5.7014121,
        6.0341221,
        5.9002733,
        5.0788391
      ),
      Difference_Gln = c(
        20.2203384,
        21.9353406,
        17.1303398,
        17.3148438,
        8.7804598,
        7.0612042,
        0.5848933,
        0.3856946,
        0.0748362,
        10.761994
      ),
      Difference_S_LDL_FC_pct = c(
        20.1171041,
        21.1194979,
        19.0084278,
        22.7941105,
        0.6624931,
        1.2516315,
        0.4276689,
        1.3559095,
        1.7227713,
        18.2480865
      ),
      Difference_Pyruvate = c(
        17.5854511,
        19.9410449,
        12.7105925,
        11.9623687,
        10.4788242,
        8.6559229,
        0.0189435,
        0.0239544,
        0.00312,
        16.7369868
      ),
      Difference_L_LDL_TG = c(
        13.250508,
        13.9865028,
        13.4782952,
        14.1116228,
        4.0569956,
        3.3975936,
        0.8144706,
        0.7677175,
        0.0572463,
        14.4629904
      ),
      Difference_S_HDL_C_pct = c(
        12.9346568,
        13.1410302,
        16.6905244,
        22.6788084,
        1.7590544,
        1.3980602,
        2.9658478,
        4.8456946,
        2.7936213,
        27.68029
      ),
      Difference_S_HDL_PL_pct = c(
        12.864684,
        11.9193723,
        10.8118128,
        11.8431789,
        1.0692856,
        1.4855054,
        1.9670692,
        1.3530411,
        8.5227632,
        -1.6739852
      ),
      Difference_M_LDL_TG_pct = c(
        12.8530009,
        11.4667269,
        13.5677286,
        16.3305456,
        0.9506285,
        0.5322617,
        0.8908791,
        1.898937,
        2.9490729,
        29.5990618
      ),
      Difference_L_LDL_C_pct = c(
        12.5509812,
        13.624991,
        12.482877,
        13.1641219,
        0.3197067,
        0.0195839,
        8.7725257,
        7.2044468,
        0.0122509,
        8.4407425
      ),
      Difference_IDL_TG = c(
        12.0014,
        11.378059,
        6.0707129,
        6.2593626,
        8.3150096,
        9.6032892,
        4.2182798,
        3.8636934,
        4.2984429,
        1.8889026
      )
    ),
    row.names = c(NA, 10L),
    class = "data.frame"
  )

# 转为热图矩阵：行=Y轴代谢物差异指标，列=X轴营养物指标
heat_mat <- data.str %>%
  column_to_rownames("Metabolite") %>%
  as.matrix() %>%
  t()

# 按实际需求修改Y轴亚部分组，示例分组如下
row_group <- c(
  rep("脂质类指标", 9),
  rep("氨基酸类指标", 2)
)
# 亚组配色可自行调整
group_col <- c(
  "脂质类指标" = "#fcbba1",
  "氨基酸类指标" = "#fb6a4a"
)

第三步：配置热图核心参数 #

完全匹配需求设置：

• 色阶范围固定为-2~45
• 红色系渐变，数值越高颜色越深
• 仅数值大于5的单元格显示对应数值

# 配置红色系色阶
col_fun <- colorRamp2(
  breaks = c(-2, 20, 45),
  colors = c("#fff5f0", "#fc9272", "#67000d")
)

# 配置单元格显示文本
cell_label <- matrix("", nrow = nrow(heat_mat), ncol = ncol(heat_mat))
cell_label[heat_mat > 5] <- round(heat_mat[heat_mat > 5], 1)

第四步：绘制热图 #

Heatmap(
  matrix = heat_mat,
  name = "差异值",
  col = col_fun,
  # 聚类开关，需要排序聚类改为TRUE即可
  cluster_rows = FALSE,
  cluster_columns = FALSE,
  # 单元格文本
  cell_fun = function(j, i, x, y, width, height, fill) {
    grid.text(cell_label[i,j], x, y, gp = gpar(fontsize = 8))
  },
  # Y轴亚组注释
  left_annotation = rowAnnotation(
    指标分类 = row_group,
    col = list(指标分类 = group_col)
  ),
  # 样式调整
  rect_gp = gpar(col = "white", lwd = 1),
  row_names_gp = gpar(fontsize = 9),
  column_names_gp = gpar(fontsize = 9),
  column_names_rot = 45
)

调整说明 #

• 替换自己的完整数据集时，保持和示例数据结构一致即可：第一列为营养物名称，其余列为代谢物差异指标，代码会自动适配规模
• Y轴亚组的分类、颜色、数值显示阈值、色阶深浅都可以直接在对应代码段修改
• 开启行/列聚类时，亚组注释会自动跟随行顺序排列，不需要额外调整

内容的提问来源于stack exchange，提问作者Ram6