TCGA临床数据整理R语言代码

由于TCGA数据集较大，且需要通过API获取，因此本回答提供的是对已经下载好的TCGA数据进行整理的R语言代码。

1. 加载所需的R包和数据

library(tidyverse) # 数据整理和可视化
library(survival) # 生存分析
library(survminer) # 生存分析可视化
library(readr) # 读取数据

# 读取临床数据
clinical_data <- read_csv("clinical_data.csv")

2. 数据清洗和整理

# 删除无用的列
clinical_data <- select(clinical_data, -c(1:4, 6, 8:10, 12:16, 18:20, 22:28, 30:34))

# 将时间格式转换为日期格式
clinical_data$date_of_initial_pathologic_diagnosis <- as.Date(clinical_data$date_of_initial_pathologic_diagnosis, "%m/%d/%Y")

# 将肿瘤分期标准化为TNM分期
clinical_data$tumor_stage <- case_when(
  str_detect(clinical_data$tumor_stage, "stage 1[a|b]") ~ "T1N0M0",
  str_detect(clinical_data$tumor_stage, "stage 2[a|b]") ~ "T2N0M0",
  str_detect(clinical_data$tumor_stage, "stage 3[a|b]") ~ "T3N0M0",
  str_detect(clinical_data$tumor_stage, "stage 4[a|b]") ~ "T4N0M0",
  str_detect(clinical_data$tumor_stage, "stage 1[c|d]") ~ "T1N1M0",
  str_detect(clinical_data$tumor_stage, "stage 2[c|d]") ~ "T2N1M0",
  str_detect(clinical_data$tumor_stage, "stage 3[c|d]") ~ "T3N1M0",
  str_detect(clinical_data$tumor_stage, "stage 4[c|d]") ~ "T4N1M0",
  TRUE ~ clinical_data$tumor_stage
)

# 将年龄分组
clinical_data$age_group <- cut(clinical_data$age_at_initial_pathologic_diagnosis, breaks = c(0, 40, 60, Inf), labels = c("<=40", "41-60", ">60"))

# 将性别和肿瘤类型转换为因子变量
clinical_data$gender <- as.factor(clinical_data$gender)
clinical_data$tumor_type <- as.factor(clinical_data$tumor_type)

# 将死亡状态转换为生存时间
clinical_data$survival_time <- case_when(
  clinical_data$vital_status == "dead" ~ clinical_data$days_to_death,
  clinical_data$vital_status == "alive" & !is.na(clinical_data$days_to_last_follow_up) ~ clinical_data$days_to_last_follow_up,
  TRUE ~ NA_integer_
)

# 删除无法计算生存时间的行
clinical_data <- clinical_data %>% filter(!is.na(survival_time))

# 计算生存状态
clinical_data$survival_status <- ifelse(clinical_data$vital_status == "dead", 1, 0)

# 删除无用的列
clinical_data <- select(clinical_data, -c(5, 7, 11, 17, 21, 29))

# 查看数据
head(clinical_data)

3. 生存分析

# 生存曲线
fit <- survfit(Surv(survival_time/365, survival_status) ~ tumor_stage, data = clinical_data)
ggsurvplot(fit, conf.int = TRUE, risk.table = TRUE)

# Cox比例风险模型
cox_model <- coxph(Surv(survival_time/365, survival_status) ~ tumor_stage + age_group + gender + tumor_type, data = clinical_data)
summary(cox_model)

以上是对TCGA临床数据的整理和生存分析的R语言代码。注：本代码仅供参考，具体应用需要根据数据情况进行适当修改。