火山图 (Volcano Plot)¶

用途：展示差异表达分析或生物标志物分析中 fold change 和 p 值的关系，快速识别统计显著且效应量大的特征。

交互式图表¶

红色 = 显著 (p<0.05 且 |log2FC|>1)，橙色 = 仅 FC 变化，蓝色 = 仅显著，灰色 = 无差异。

生成代码¶

RPython

library(plotly)

# 模拟数据
set.seed(123)
volcano_data <- data.frame(
  log2FC = rnorm(1000, 0, 1.5),
  pval = runif(1000, 0, 1)
)
volcano_data$logp <- -log10(volcano_data$pval)
volcano_data$sig <- ifelse(
  abs(volcano_data$log2FC) > 1 & volcano_data$pval < 0.05, "Significant",
  ifelse(abs(volcano_data$log2FC) > 1, "FC only",
         ifelse(volcano_data$pval < 0.05, "p only", "NS"))
)

fig <- plot_ly(volcano_data, 
               x = ~log2FC, y = ~logp,
               color = ~sig,
               colors = c("Significant" = "#d62728", "FC only" = "#ff7f0e",
                          "p only" = "#1f77b4", "NS" = "gray"),
               type = "scatter", mode = "markers",
               alpha = 0.6, size = 2,
               hovertemplate = "FC: %{x:.2f}<br>-log10(p): %{y:.2f}<extra></extra>") |>
  layout(
    title = "Volcano Plot",
    xaxis = list(title = "log2(Fold Change)"),
    yaxis = list(title = "-log10(p value)"),
    shapes = list(
      list(type = "line", x0 = -1, x1 = -1, y0 = 0, y1 = max(volcano_data$logp),
           line = list(dash = "dash", color = "gray")),
      list(type = "line", x0 = 1, x1 = 1, y0 = 0, y1 = max(volcano_data$logp),
           line = list(dash = "dash", color = "gray"))
    )
  )
fig

import plotly.express as px
import pandas as pd
import numpy as np

np.random.seed(123)
volcano_data = pd.DataFrame({
    "log2FC": np.random.normal(0, 1.5, 1000),
    "pval": np.random.uniform(0, 1, 1000)
})
volcano_data["logp"] = -np.log10(volcano_data["pval"])

conditions = [
    (abs(volcano_data["log2FC"]) > 1) & (volcano_data["pval"] < 0.05),
    (abs(volcano_data["log2FC"]) > 1),
    (volcano_data["pval"] < 0.05)
]
choices = ["Significant", "FC only", "p only"]
volcano_data["sig"] = np.select(conditions, choices, default="NS")

fig = px.scatter(volcano_data, x="log2FC", y="logp", color="sig",
                 color_discrete_map={
                     "Significant": "#d62728", "FC only": "#ff7f0e",
                     "p only": "#1f77b4", "NS": "gray"},
                 opacity=0.6,
                 title="Volcano Plot")
fig.add_vline(x=-1, line_dash="dash", line_color="gray")
fig.add_vline(x=1, line_dash="dash", line_color="gray")
fig.show()

关键参数¶

x 轴: log2(Fold Change)，通常阈值 = ±1 (即 ±2 倍变化)
y 轴: -log10(p value)，通常阈值 = 1.3 (即 p < 0.05)

参考方法¶

多重比较校正