Time ROC

Note

Hiplot website

This page is the tutorial for source code version of the Hiplot Time ROC plugin. You can also use the Hiplot website to achieve no code ploting. For more information please see the following link:

https://hiplot.cn/basic/time-roc?lang=en

Receiver Operating Characteristic (ROC) analysis with time records in survival analysis.

Setup

• System Requirements: Cross-platform (Linux/MacOS/Windows)

• Programming language: R

• Dependent packages: plotROC; survivalROC; ggplot2; grid

# Install packages
if (!requireNamespace("plotROC", quietly = TRUE)) {
  install.packages("plotROC")
}
if (!requireNamespace("survivalROC", quietly = TRUE)) {
  install.packages("survivalROC")
}
if (!requireNamespace("ggplot2", quietly = TRUE)) {
  install.packages("ggplot2")
}
if (!requireNamespace("grid", quietly = TRUE)) {
  install.packages("grid")
}

# Load packages
library(plotROC)
library(survivalROC)
library(ggplot2)
library(grid)

Data Preparation

• : (Numeric) survival data (i.e survive, risk).
• : (Numeric) time data.

# Load data
data1 <- read.delim("files/Hiplot/171-time-roc-data1.txt", header = T)
data2 <- read.delim("files/Hiplot/171-time-roc-data2.txt", header = T)

# convert data structure
surv_table <- data1
colnames(surv_table) <- c("surv", "cens", "risk")
mtime <- as.data.frame(data2)[, 1]
sroc <- lapply(mtime, function(t) {
  stroc <- survivalROC(
    Stime = surv_table$surv,
    status = surv_table$cens,
    marker = surv_table$risk,
    predict.time = t,
    method = "KM"
  )
  data.frame(
    TPF = stroc[["TP"]],
    FPF = stroc[["FP"]],
    cut = stroc[["cut.values"]],
    time = rep(
      stroc[["predict.time"]],
      length(stroc[["TP"]])
    ),
    AUC = rep(
      stroc$AUC,
      length(stroc$FP)
    )
  )
})
mroc <- do.call(rbind, sroc)
mroc$time <- factor(mroc$time)

# View data
head(data1)

       surv cens       risk
1 11.126027    0 0.19205450
2  9.794521    0 0.47734974
3 13.690411    0 0.04605343
4 10.068493    0 0.29717146
5  3.317808    0 0.18144610
6 12.312329    0 0.62681895

head(data2)

  times
1     2
2     4
3     6
4     8
5    10

Visualization

# Time ROC
col <- c("#E64B35FF","#4DBBD5FF","#00A087FF","#3C5488FF","#F39B7FFF")
p <- ggplot(mroc, aes(x = FPF, y = TPF, label = cut, color = time)) +
  plotROC::geom_roc(labels = FALSE, stat = "identity", n.cuts = 0) +
  geom_abline(slope = 1, intercept = 0, color = "red", linetype = 2) +
  labs(title = "ROC Dependence Time", x = "False positive rate",
       y = "True positive rate", 
       color = paste("Time", "(", "Year", ")")) +
  theme_bw() +
  theme(text = element_text(family = "Arial"),
        plot.title = element_text(size = 12, hjust = 0.5),
        axis.title = element_text(size = 10),
        legend.position = "right",
        legend.direction = "vertical",
        legend.title = element_text(size = 10),
        legend.text = element_text(size = 10)) +
  scale_color_manual(values = col)

auc <- levels(factor(mroc$AUC))
for (i in 1:length(auc)) {
  p <- p + annotate("text",
    x = 0.75,
    y = 0.05 + 0.05 * i, ## 注释text的位置
    col = col[i],
    label = paste(
      paste(paste(mtime[i], "Year", sep = " "), " = "),
      round(as.numeric(auc[i]), 2)
    )
  )
}

p

Figure 1: Time ROC