kaplan_meier_plot.qmd

---
title: "kaplan_meier_plot"
format: gfm
editor: visual
execute:
  echo: true
  warning: false
  message: false

---
```{r setup, include=FALSE}
# Load required libraries-----
library(tidyverse)
library(survival)
library(survminer)
library(dplyr)
```

```{r}
#Load Data-----
surv_data <- read_tsv("survival_data.tsv") #There are 60 TARGET samples with survival data"
clusters <- read_tsv("cluster_assignments_targetsamples.tsv") 
```

```{r}
#Data clean up-----
#There are 75 TARGET samples with cluster assignements
clusters <- clusters %>% rename(cluster_id = Cluster)

# Filter cluster data to match samples present in survival data
common_samples <- intersect(surv_data$sample, clusters$sample)

surv_data_filtered <- surv_data %>% filter(sample %in% common_samples)
clusters_filtered <- clusters %>% filter(sample %in% common_samples)

# Merge the data by sample ID
merged_data <- inner_join(surv_data_filtered, clusters_filtered, by = "sample")

```

```{r}
# Data Check-----
# Check if you have survival status data (e.g., OS.event).)
if (!"OS.event" %in% colnames(merged_data)) {
  merged_data$OS.event <- 1  # Assume all events occurred (placeholder)
}
```

```{r}
#Data Preparation for plotting-----

# Convert cluster_id to factor for plotting
merged_data$cluster_id <- as.factor(merged_data$cluster_id)
table(merged_data$OS.event, useNA = "ifany")

surv_object <- Surv(time = merged_data$OS.time, event = merged_data$OS.event)
```

```{r}
#Designing Fit Survival Curves 

fit <- survfit(surv_object ~ cluster_id, data = merged_data)

summary(merged_data$OS.time)
summary(merged_data$OS.event)
summary(merged_data$cluster_id)
```

```{r}
# Plot Kaplan-Meier curves-----
ggsurvplot(fit, data = merged_data,
           pval = TRUE,
           risk.table = TRUE,
           surv.median.line = "hv",
           ggtheme = theme_minimal(),
           palette = "Dark2")
```

```{r}
# Save the KM plot and Export Image-----
km_plot <- ggsurvplot(fit, data = merged_data,
                      pval = TRUE,
                      risk.table = TRUE,
                      surv.median.line = "hv",
                      ggtheme = theme_minimal(),
                      palette = "Dark2")

# Export the plot (you can set file type by extension)
ggsave("kaplan_meier_plot_target_assignments.png", plot = km_plot$plot, width = 8, height = 6, dpi = 300)
```

```{r}
#Log-Test-----

# Run the log-rank test to compare survival curves across clusters
log_rank_test <- survdiff(surv_object ~ cluster_id, data = merged_data)

# Display the results of the log-rank test
log_rank_test
```