求R代码绘制昼夜节律数据环形箱线图

需求概述 #

• 复刻目标：某PubMed文献中的环形箱线图（原Python绘制，无公开代码）
• 数据结构：包含sample（BAT、WAT、in vitro三类）、process（昼夜节律调控检测指标）、ct（24小时时间点，0/24点无缝衔接）
• 可视化要求：为每个sample生成独立图表，展示各process的相位分布，需包含中位数与IQR（四分位距）

数据示例 #

sample   process            ct
    BAT CellCycle   1.140128762
    BAT CellCycle   3.106516768
    BAT CellCycle   3.738113578
    BAT CellCycle   23.44912305
    BAT CellCycle   23.49744379
    BAT CellCycle   23.51422441
    BAT CellCycle   23.33041125
    BAT CellCycle   22.19656322
    BAT CellCycle   22.84183477
    BAT ECM 22.04149961
    BAT ECM 23.45526633
    BAT ECM 23.36802574
    BAT ECM 23.40688727
    BAT ECM 1.732071528
    BAT ECM 1.529777828
    BAT ECM 20.29377917
    BAT Energy  2.19647636
    BAT Energy  4.177766894
    BAT Energy  2.624942744
    BAT Energy  1.633080723
    BAT Energy  2.25815373
    BAT Energy  22.52527935
    BAT Energy  21.844178
    BAT Energy  1.44484992
    BAT Energy  1.710318176
    BAT Energy  1.721398481
    BAT Energy  5.0700794
    BAT Energy  3.252122282
    BAT Energy  19.69284419
    BAT Energy  2.521538024
    BAT Energy  2.309621407
    BAT Energy  1.979057875
    BAT Energy  1.747326389
    BAT Energy  1.620003554
    BAT Energy  4.512535826
    BAT Energy  19.48505391
    BAT Energy  3.96613478

R实现代码 #

基于tidyverse和ggplot2，处理0/24点衔接问题，生成环形箱线图：

# 加载依赖包
library(tidyverse)

# 替换为你的完整数据集
df <- tibble(
  sample = rep("BAT", 38),
  process = c(rep("CellCycle",9), rep("ECM",7), rep("Energy",22)),
  ct = c(1.140128762,3.106516768,3.738113578,23.44912305,23.49744379,23.51422441,23.33041125,22.19656322,22.84183477,
         22.04149961,23.45526633,23.36802574,23.40688727,1.732071528,1.529777828,20.29377917,
         2.19647636,4.177766894,2.624942744,1.633080723,2.25815373,22.52527935,21.844178,1.44484992,1.710318176,1.721398481,5.0700794,3.252122282,19.69284419,2.521538024,2.309621407,1.979057875,1.747326389,1.620003554,4.512535826,19.48505391,3.96613478)
)

# 处理时间点：转换为极坐标角度，同时解决0/24衔接问题
df <- df %>%
  mutate(ct_wrap = ifelse(ct > 22, ct - 24, ct))  # 将22点后的数据转换为负数值，实现环形衔接

# 按sample分组生成并保存图表
df %>%
  group_by(sample) %>%
  group_walk(function(data, group) {
    plot <- ggplot(data, aes(x = process, y = ct_wrap)) +
      geom_boxplot(width = 0.8, fill = "#69b3a2", alpha = 0.7) +
      coord_polar(start = -pi/2, direction = 1) +  # 极坐标：0点对应顶部，顺时针旋转
      scale_y_continuous(
        breaks = c(-2, 0, 2, 4, 6, 8, 10, 12, 14, 16, 18, 20, 22),
        labels = c("22", "0", "2", "4", "6", "8", "10", "12", "14", "16", "18", "20", "22")
      ) +
      theme_minimal() +
      theme(
        panel.grid.major.y = element_line(color = "gray80"),
        panel.grid.major.x = element_blank(),
        axis.title = element_blank(),
        plot.title = element_text(hjust = 0.5, size = 16),
        axis.text.y = element_text(size = 10),
        axis.text.x = element_text(size = 12, face = "bold")
      ) +
      ggtitle(paste("相位分布 -", group$sample))
    
    ggsave(paste0("环形箱线图_", group$sample, ".png"), plot, width = 8, height = 8, dpi = 300)
    print(plot)
  })

Python实现代码 #

基于matplotlib和seaborn，实现同效果的环形箱线图：

import pandas as pd
import matplotlib.pyplot as plt
import seaborn as sns
import numpy as np

# 替换为你的完整数据集
data = {
    'sample': ['BAT']*38,
    'process': ['CellCycle']*9 + ['ECM']*7 + ['Energy']*22,
    'ct': [1.140128762,3.106516768,3.738113578,23.44912305,23.49744379,23.51422441,23.33041125,22.19656322,22.84183477,
           22.04149961,23.45526633,23.36802574,23.40688727,1.732071528,1.529777828,20.29377917,
           2.19647636,4.177766894,2.624942744,1.633080723,2.25815373,22.52527935,21.844178,1.44484992,1.710318176,1.721398481,5.0700794,3.252122282,19.69284419,2.521538024,2.309621407,1.979057875,1.747326389,1.620003554,4.512535826,19.48505391,3.96613478]
}
df = pd.DataFrame(data)

# 处理时间点，解决0/24衔接问题
df['ct_wrap'] = df['ct'].apply(lambda x: x - 24 if x > 22 else x)

# 按sample分组生成图表
for sample_name, group_data in df.groupby('sample'):
    fig, ax = plt.subplots(figsize=(8,8), subplot_kw={'projection': 'polar'})
    
    # 映射process到极坐标位置
    process_list = group_data['process'].unique()
    process_indices = np.arange(len(process_list))
    process_to_idx = {p:i for i,p in enumerate(process_list)}
    group_data['pos'] = group_data['process'].map(process_to_idx)
    
    # 绘制箱线图
    sns.boxplot(
        data=group_data, x='pos', y='ct_wrap',
        ax=ax, width=0.8, palette='viridis', alpha=0.7
    )
    
    # 极坐标配置：0点对应顶部，顺时针旋转
    ax.set_theta_zero_location('N')
    ax.set_theta_direction(1)
    
    # 设置y轴刻度与标签（24小时制）
    ax.set_yticks([-2,0,2,4,6,8,10,12,14,16,18,20,22])
    ax.set_yticklabels(['22','0','2','4','6','8','10','12','14','16','18','20','22'])
    
    # 设置x轴刻度为process名称
    ax.set_xticks(process_indices * 2 * np.pi / len(process_list))
    ax.set_xticklabels(process_list, fontsize=12, fontweight='bold')
    
    # 调整样式
    ax.grid(True, axis='y', color='gray')
    ax.set_title(f'相位分布 - {sample_name}', y=1.1, fontsize=16)
    
    # 保存并展示
    plt.savefig(f'环形箱线图_{sample_name}.png', dpi=300, bbox_inches='tight')
    plt.show()

内容的提问来源于stack exchange，提问作者Jen1984