生信分析（2）——代谢物热图（附代码）

代谢物热图绘制

代谢物热图说明

如图是一张代谢物热图（heatmap），常用于展示不同样本之间代谢物相对丰度的变化趋势和聚类关系，关于图的说明如下：

横轴（列）：表示不同的样本，每一列是一个样本。

顶部颜色条标注了分组：

红色（pt-DVT）：患者组

蓝绿色（Controls）：对照组

纵轴（行）：表示不同的代谢物，每一行为一个代谢物。

右侧列出代谢物名称。

颜色强度表示代谢物在样本中的表达水平（强度）：

红色：上调（表达增加）

蓝色：下调（表达减少）

白色：中间值（无显著差异）

左侧树状图（dendrogram）：是层次聚类分析（hierarchical clustering）结果，表示代谢物在不同样本中表达模式的相似性。

左侧的色条表示每个代谢物的分类（Class）：

如绿色：Amino（氨基酸类）、橙色：Carbohydrate（碳水化合物类）等

右下角的图例说明：

Group：样本分组

Significant：显著性变化方向

Class：代谢物的功能分类

绘制流程

1.数据准备

Compounds	Class I	Type	样本1	样本2	···
代谢物1	代谢物类别	up	123	456	···
代谢物2	代谢物类别	down	123	456	···
···	···	···	···	···	···

2.绘制代码

library(ComplexHeatmap)
library(circlize)
library(RColorBrewer)

# ========= 读取数据 =========
data <- read.csv("你的路径/你的文件.csv", check.names = FALSE)
# 4代表从第四列开始为样本数据
expr_raw <- data[, 4:ncol(data)]
rownames(expr_raw) <- data$Compound

# ========= 标准化 =========
expr_norm <- sweep(expr_raw, 2, colSums(expr_raw), FUN = "/")
expr_log <- log10(expr_norm + 1)
expr_scaled <- t(scale(t(expr_log)))

# ========= 样本注释 =========
sample_names <- colnames(expr_scaled)
# 这里要修改140这2个值，140代表从第1个样本到第140个样本为Cancer样本，140之后的为健康样本
# 注意是从第1个样本到第140个样本，不是从第1列到第140列
group_info <- data.frame(Group = c(rep("Cancer", 140), rep("Control", length(sample_names) - 140)))
rownames(group_info) <- sample_names

# ========= 代谢物注释 =========
row_anno <- data.frame(
  Significant = data$Type,
  Class = data$`Class I`
)
rownames(row_anno) <- data$Compound

# ========= 自定义颜色 =========
col_fun <- colorRamp2(c(-4, 0, 4), c("#4575B4", "white", "#D73027"))
group_col <- c("Cancer" = "#EF8A62", "Control" = "#67A9CF")
sig_col <- c("up" = "#EF8A9E", "down" = "#9ECAE1")
class_levels <- unique(row_anno$Class)
class_col <- setNames(brewer.pal(n = length(class_levels), "Set3"), class_levels)

# ========= 顶部注释（仅颜色条，无图例）=========
top_anno <- HeatmapAnnotation(
  Group = group_info$Group,
  col = list(Group = group_col),
  show_annotation_name = TRUE,               # ✅ 显示标签
  annotation_name_side = "right",             # ✅ 标签在颜色条左侧
  show_legend = FALSE
)

# ========= 行侧注释（显示颜色块，无图例）=========
left_anno <- rowAnnotation(
  Class = row_anno$Class,
  Significant = row_anno$Significant,
  col = list(Class = class_col, Significant = sig_col),
  show_annotation_name = FALSE,
  show_legend = FALSE
)

# ========= 热图主体 =========
ht <- Heatmap(expr_scaled,
              name = "Intensity",
              col = col_fun,
              show_row_names = TRUE,
              row_names_side = "right",
              row_names_gp = gpar(fontsize = 7),
              show_column_names = FALSE,
              cluster_rows = TRUE,
              cluster_columns = FALSE,
              top_annotation = top_anno,
              left_annotation = left_anno,
              show_heatmap_legend = FALSE,
              column_title = "Differential Metabolites Expression Heatmap",
              column_title_gp = gpar(fontsize = 14, fontface = "bold"),
              heatmap_legend_param = list(
                title = "Intensity",
                title_gp = gpar(fontsize = 12, fontface = "bold"),
                labels_gp = gpar(fontsize = 10),
                at = c(-4, -2, 0, 2, 4),
                direction = "vertical"
              ))

# ========= 图例打包，显示在右侧 =========
# 提取主图例（Intensity）
lgd_intensity <- Legend(
  title = "Intensity",
  at = c(-4, -2, 0, 2, 4),
  col_fun = col_fun,
  direction = "vertical",
  title_gp = gpar(fontsize = 12, fontface = "bold"),
  labels_gp = gpar(fontsize = 10)
)

# 打包所有图例，Intensity 放最上面
packed_lgd <- packLegend(
  list = list(
    Legend(title = "Group", at = names(group_col), legend_gp = gpar(fill = group_col), 
           title_gp = gpar(fontface = "bold", fontsize = 10), labels_gp = gpar(fontsize = 9)),
    lgd_intensity,
    Legend(title = "Significant", at = names(sig_col), legend_gp = gpar(fill = sig_col), 
           title_gp = gpar(fontface = "bold", fontsize = 10), labels_gp = gpar(fontsize = 9)),
    Legend(title = "Class", at = names(class_col), legend_gp = gpar(fill = class_col), 
           title_gp = gpar(fontface = "bold", fontsize = 10), labels_gp = gpar(fontsize = 8), ncol = 1)
  ),
  direction = "vertical",
  gap = unit(4, "mm")  # 控制图例之间间距
)

# 绘图
draw(ht,
     annotation_legend = packed_lgd,
     annotation_legend_side = "right")