医学科研绘图：年底了，我还在做火山图返修论文，圣诞老人当场对我敬礼！

1. 前言

我时不时的会收到后台私信，想看看火山图的化教程。我对火山图也很高兴去，医学数据的基因结果可视化常常用到。众所周知，火山图（Volcano Plot）可以同时展示效应大小和统计显著性，直观地揭示差异显著的基因、蛋白质或代谢物。火山图不仅在基因表达分析中应用广泛，还在蛋白质组学、代谢组学、GWAS 和药物研究中使用。

今天的更新我们将介绍火山图的概念，并结合 R 语言代码演示其在医学数据中的具体应用。

2. 医学数据中的应用与代码示例

2.1 差异基因表达分析

情境：比较病例组和对照组的基因表达水平，寻找显著上调或下调的基因。

# 加载必要的库
library(ggplot2)
library(ggrepel)

# 模拟数据
set.seed(123)
n <- 1000

# 非显著基因
non_sig <- data.frame(
  Gene = paste0("Gene", 1:(n * 0.7)),
  log2FC = rnorm(n * 0.7, mean = 0, sd = 0.5),
  pvalue = runif(n * 0.7, min = 0.05, max = 1)
)

# 显著上调基因
upregulated <- data.frame(
  Gene = paste0("Gene_up", 1:(n * 0.15)),
  log2FC = rnorm(n * 0.15, mean = 2, sd = 0.5),
  pvalue = runif(n * 0.15, min = 0, max = 0.01)
)

# 显著下调基因
downregulated <- data.frame(
  Gene = paste0("Gene_down", 1:(n * 0.15)),
  log2FC = rnorm(n * 0.15, mean = -2, sd = 0.5),
  pvalue = runif(n * 0.15, min = 0, max = 0.01)
)

# 合并数据
gene_data <- rbind(non_sig, upregulated, downregulated)

# 添加显著性标签
gene_data$Significance <- with(gene_data, ifelse(
  pvalue < 0.01 & abs(log2FC) > 1, 
  ifelse(log2FC > 0, "Upregulated", "Downregulated"), 
  "Not Significant"
))

# 绘制优化后的火山图
ggplot(gene_data, aes(x = log2FC, y = -log10(pvalue), color = Significance)) +
  geom_point(alpha = 0.8, size = 2) +
  scale_color_manual(values = c("Upregulated" = "#E41A1C", 
                                "Downregulated" = "#377EB8", 
                                "Not Significant" = "grey80")) +
  theme_minimal(base_size = 15) +
  labs(
    title = "Differential Gene Expression Volcano Plot",
    x = expression(log[2]~Fold~Change),
    y = expression(-log[10]~p~value)
  ) +
  theme(
    legend.position = "top",
    plot.title = element_text(hjust = 0.5, size = 18, face = "bold"),
    axis.title = element_text(size = 14),
    panel.grid.minor = element_blank()
  ) +
  geom_text_repel(
    data = subset(gene_data, Significance != "Not Significant"),
    aes(label = Gene),
    size = 3,
    max.overlaps = 10
  )

2.2 代谢组学研究

情境：分析代谢产物在不同组间的变化。

# 加载必要的库
library(ggplot2)
library(ggrepel)

# 模拟数据
set.seed(456)
metabolite_data <- data.frame(
  Metabolite = paste0("Metabolite", 1:500),
  log2FC = rnorm(500, mean = 0, sd = 1.5),
  pvalue = runif(500, min = 0, max = 0.05)
)

# 添加显著性标签
metabolite_data$Significance <- with(metabolite_data, ifelse(
  pvalue < 0.01 & abs(log2FC) > 1, 
  ifelse(log2FC > 0, "Increased", "Decreased"), 
  "Not Significant"
))

# 绘制火山图
ggplot(metabolite_data, aes(x = log2FC, y = -log10(pvalue), color = Significance)) +
  geom_point(alpha = 0.9, size = 2.5) +
  scale_color_manual(values = c("Increased" = "#FBB4AE", 
                                "Decreased" = "#B3CDE3", 
                                "Not Significant" = "grey70")) +
  theme_classic(base_size = 15) +
  labs(
    title = "Metabolomics Volcano Plot",
    x = expression(log[2]~Fold~Change),
    y = expression(-log[10]~p~value)
  ) +
  theme(
    legend.position = "bottom",
    plot.title = element_text(hjust = 0.5, size = 18, face = "bold"),
    panel.background = element_rect(fill = "#F0F0F0"),
    axis.line = element_line(size = 0.8, color = "black")
  ) +
  geom_text_repel(
    data = subset(metabolite_data, Significance != "Not Significant"),
    aes(label = Metabolite),
    size = 3,
    max.overlaps = 10
  )

2.3 GWAS（基因组关联研究）

情境：筛选与疾病相关的 SNP。

# 加载必要的库
library(ggplot2)
library(ggrepel)

# 模拟数据
set.seed(789)
snp_data <- data.frame(
  SNP = paste0("rs", sample(1:10000, 1000)),
  log2OR = rnorm(1000, mean = 0, sd = 1),
  pvalue = runif(1000, min = 0, max = 0.05)
)

# 添加显著性标签
snp_data$Significance <- with(snp_data, ifelse(
  pvalue < 0.001 & abs(log2OR) > 0.8, 
  ifelse(log2OR > 0, "Risk", "Protective"), 
  "Not Significant"
))

# 绘制火山图
ggplot(snp_data, aes(x = log2OR, y = -log10(pvalue), color = -log10(pvalue))) +
  geom_point(alpha = 0.9, size = 2) +
  scale_color_gradientn(colors = c("#0571B0", "#92C5DE", "#F4A582", "#CA0020")) +
  theme_minimal(base_size = 15) +
  labs(
    title = "GWAS Volcano Plot",
    x = expression(log[2]~Odds~Ratio),
    y = expression(-log[10]~p~value),
    color = expression(-log[10]~p~value)
  ) +
  theme(
    legend.position = "right",
    plot.title = element_text(hjust = 0.5, size = 18, face = "bold"),
    panel.grid.major = element_line(color = "grey80", size = 0.5)
  ) +
  geom_text_repel(
    data = subset(snp_data, Significance != "Not Significant"),
    aes(label = SNP),
    size = 3,
    max.overlaps = 10
  )

3. 小结

今天的更新，我们展示了如何使用 R 语言生成美学优化后的火山图。因为我使用的都是模拟数据，所以效果可能不是很直观。但是这些代码的参数是通用的，相信你使用你自己的数据，再结合我的代码，可以绘制出风格简洁、信息清晰，能够满足高质量科研要求的火山图。如果有任何问题，欢迎留言交流。

感谢关注，你的支持是我不懈的动力！