Skip to content

Latest commit

 

History

History
128 lines (108 loc) · 5.11 KB

Read-multiple-file_reshape_ggplot.md

File metadata and controls

128 lines (108 loc) · 5.11 KB
title author date output
Read multiple files, reshape and plot
Quynh Nhu Nguyen
2023-04-09
html_document

READ MULTIPLE FILES, RESHAPE AND PLOT

# Load library 
library(tidyverse)
library(data.table)
library(reshape) # reshape a data frame to 'wide' format

# Set the root directory for notebook chunks
knitr::opts_knit$set(root.dir = "C:/Users/lucia/Downloads/Lecturer from Dr. Loi/NGS3/exam_data/exam_data")

1. Read file “tools_overlap_filter_1.csv” to “tools_overlap_filter_11.csv” into R and merge them into a single data frame.

# Create an empty list to store the data frames
tools_overlap_filter_list <- list()

# Loop over the file names and read them into the list
for (i in 1:11) {
  file_name <- paste0("tools_overlap_filter_", i, ".csv")
  df <- read.csv(file_name)
  
  # Add a case column with the current iteration number
  df$Cases <- paste0("Case_", i)
  
  # Add the data frame to the list
  tools_overlap_filter_list[[i]] <- df
}

# Merge the data frames into a single data frame
tools_overlap_filter_combine <- do.call(rbind, tools_overlap_filter_list)

In R, the double square bracket notation [[ is used to extract or replace elements of a list by their index or name. It is different from the single square bracket notation [ in that it returns the actual element of the list, rather than a sub-list containing that element.

In the line tools_overlap_filter_list[[i]] <- df, the double square brackets are used to set thei-th element of the tools_overlap_filter_list list to be the data frame df. The double brackets are used instead of single brackets because we want to set the actual element of the list to be the data frame, rather than creating a sub-list containing the data frame.

# Check data
head(tools_overlap_filter_combine)
dim(tools_overlap_filter_combine)
str(tools_overlap_filter_combine)

# Check and remove NA 
colSums(is.na(tools_overlap_filter_combine))
tools_overlap_filter_combine <- na.omit(tools_overlap_filter_combine)
head(tools_overlap_filter_combine)

# Move the "Cases" column to 1st column
tools_overlap_filter_combine <- tools_overlap_filter_combine[, c(5, 1:4)]

# Reshape the data frame into a "long" format with separate columns for the tool and its values
data_melt_tools_overlap_filter<- reshape2::melt(tools_overlap_filter_combine)

# Rename the first column
names(data_melt_tools_overlap_filter)[2] <- "Tools"
names(data_melt_tools_overlap_filter)[3] <- "SNPs"

# View the melt data frame
data_melt_tools_overlap_filter

2. Compute and plot the mean and median of number of overlapped SNPs of Mutect2/FreeBayes, MuTect2/Strelka, FreeBayes/Mutect2/Strelka.

data_summary_tools_overlap_filter <- data_melt_tools_overlap_filter %>%
  group_by(Tools) %>%
  summarize(mean = mean(SNPs), median = median(SNPs)) %>%
  as.data.frame()
  
# View the statistical summary data
data_summary_tools_overlap_filter

# Reshape a data frame to 'wide' format
data_melt_summary_tools_overlap_filter <- reshape2::melt(data_summary_tools_overlap_filter)

# View the melt data frame
data_melt_summary_tools_overlap_filter

# Create a boxplot with the mean and median values for each tools
data_melt_summary_tools_overlap_filter %>% 
  ggplot(aes(x=Tools, y=value, fill=variable)) +
  geom_bar(stat="identity", position=position_dodge()) +
  labs(title="Mean and Median number of SNPs identified by different Tools", x="Tools", y="Number of overlapped SNPs", fill = "Variable") +
  theme(plot.title = element_text(size = 12, face = "bold")) +
  theme_classic()

3. Plot barplot with X axis is Cases and Y axis is number of overlapped SNPs of FreeBayes/Mutect2/Strelka.

# Create a barplot with X axis is Cases and Y axis is number of SNPs
data_melt_tools_overlap_filter %>%
  filter(Tools == "FreeBayes.Mutect2.Strelka") %>%
  mutate(Number = as.numeric(gsub("Case_", "", Cases))) %>%
  mutate(Cases = reorder(Cases, Number)) %>% # reorder
  ggplot(aes(x=Cases, y=SNPs, fill=Cases)) +
  geom_bar(stat="identity") +
  labs(title="Number of SNPs identified by FreeBayes.Mutect2.Strelka", x="Cases", y="Number of SNPs") +
  theme(plot.title = element_text(size = 12, face = "bold")) +
  theme_classic()

The stat parameter specifies the statistical transformation to apply to the data before plotting. In the case of geom_bar(), stat="identity" means that the height of each bar in the chart represents the actual count or value of the data, rather than any statistical summary such as the mean or median.

4. Plot boxplot with X axis is Mutect2/FreeBayes, Mutect2/Strelka, FreeBayes/Mutect2/Strelka and Y axis is number of overlapped SNPs.

# Create a boxplot for each tools
data_melt_tools_overlap_filter %>%
  filter(Tools != "FreeBayes.Strelka") %>%
  mutate(Tools = reorder(Tools, SNPs, fun = "median")) %>% # reorder
  ggplot(aes(x=Tools, y=SNPs, fill=Tools)) +
  geom_boxplot(alpha=1, na.rm = TRUE) +
  labs(title="Number of overlapped SNPs identified by different Tools", x="Tools", y="Number of overlapped SNPs") +
  theme(plot.title = element_text(size = 12, face = "bold")) +
  theme_bw()