Bioinformatics Visualization in R - Part 1

Published

July 12, 2024

Lecture 1: Introduction to Visualization and Basic Plots

Objective: - Introduce basic bioinformatics visualization techniques using R Histogram, Scatter Plot, Box Plot, Line Plot
- Overview of syntax using simulated data - Explore basic plots using nhanes data. - Exercise 1

Load necessary libraries

library(ggplot2)
library(dplyr)

Warning: package 'dplyr' was built under R version 4.3.2


Attaching package: 'dplyr'

The following objects are masked from 'package:stats':

    filter, lag

The following objects are masked from 'package:base':

    intersect, setdiff, setequal, union

library(readr)

Warning: package 'readr' was built under R version 4.3.2

library(tidyr)

Warning: package 'tidyr' was built under R version 4.3.2

Generate synthetic data

# set.seed ensures that the sequence of random numbers generated can be reproduced.
set.seed(123)

data <- data.frame(
  Time = 1:50,
  # rnorm() = random normal is used to generate random numbers from a normal distribution."
  X = rnorm(50, mean = 5, sd = 2),
  Y = 2 * rnorm(50, mean = 5, sd = 2) + rnorm(50),
  # c("A", "B") creates a vector containing the elements "A" and "B"
  # each = 25 specifies that each element in the vector should be repeated 25 times
  Group = rep(c("A", "B"), each = 25)
)

Histogram Plot: Histograms are useful to show the distribution of a single continuous variable

# Create the histogram using ggplot2
histogram_plot <- ggplot(data, aes(x = X)) +
                  geom_histogram(binwidth = (max(data$X) - min(data$X)) / 10, 
                  fill = "purple", color = "white") +
                  ggtitle("Histogram") +
                  xlab("Value") +
                  theme_minimal() +
                  theme(
                    plot.background = element_rect(fill = "white"),
                    panel.background = element_rect(fill = "white")
                  )

# Display the plot
histogram_plot

# Save the plot to a file
ggsave("3_histogram_plot.png", plot = histogram_plot, width = 8, height = 6, dpi = 300)

Scatter Plot: Scatter plots are useful to show the relationship between two continuous variables

# Create the scatter plot with a white background
Scatter_plot <- ggplot(data, aes(x = X, y = Y)) +
                geom_point() +
                ggtitle("Scatter Plot") +
                xlab("X-axis") +
                ylab("Y-axis") 

# Display the plot
Scatter_plot

# Save the plot to a file
ggsave("1_scatter_plot.png", plot = Scatter_plot, width = 8, height = 6, dpi = 300)

Box Plot: Box plots are useful to show the distribution of a continuous variable

# Box Plot
# Create the box plot
Box_plot <- ggplot(data, aes(x = Group, y = Y, fill = Group)) +
            geom_boxplot() +
            ggtitle("Box Plot") +
            xlab("Group") +
            ylab("Value") +
            scale_fill_manual(values = c("red", "green")) +
            theme_minimal() +
            theme(
              plot.background = element_rect(fill = "white"),
              panel.background = element_rect(fill = "white")
            )

# Display the plot
Box_plot

# Save the plot to a file
ggsave("2_box_plot.png", plot = Box_plot, width = 8, height = 6, dpi = 300)

Line Plot: Line plots are useful to show trends over time or another continuous variable

# Create the line plot using ggplot2
data$Value <- cumsum(rnorm(50))
line_plot <- ggplot(data, aes(x = Time, y = Value)) +
  geom_line(color = "orange", linewidth = 1.5) + # Set line color and thickness
  ggtitle("Line Plot") +
  xlab("Time") +
  ylab("Value") +
  theme_minimal() +
  theme(
    plot.background = element_rect(fill = "white"),
    panel.background = element_rect(fill = "white")
  )

# Display the plot
print(line_plot)

# Save the plot to a file
ggsave("4_line_plot.png", plot = line_plot, width = 8, height = 6, dpi = 300)

Load a healthcare dataset: nhanes (National Health and Nutrition Examination Survey)

# Actual Dataset Read the data from the provided URL
NHANES <- read_csv("https://raw.githubusercontent.com/GTPB/PSLS20/master/data/NHANES.csv", show_col_types = FALSE)

# Extract only the columns ID, Age, BMI, and BloodPressure
nhanes_subset <- NHANES %>%
  select(ID, Age, BMI, BloodPressure = BPSysAve) %>%
  distinct(ID, .keep_all = TRUE) %>%
  drop_na()

# Display the first few rows of the subset
print(head(nhanes_subset))

# A tibble: 6 × 4
     ID   Age   BMI BloodPressure
  <dbl> <dbl> <dbl>         <dbl>
1 51624    34  32.2           113
2 51630    49  30.6           112
3 51638     9  16.8            86
4 51646     8  20.6           107
5 51647    45  27.2           118
6 51654    66  23.7           111

# Write the cleaned dataset to a CSV file
write_csv(nhanes_subset, "health_data.csv")

# Display summary of the cleaned dataset
print(summary(nhanes_subset))

       ID             Age            BMI        BloodPressure
 Min.   :51624   Min.   : 8.0   Min.   :12.89   Min.   : 76  
 1st Qu.:56548   1st Qu.:22.0   1st Qu.:22.65   1st Qu.:106  
 Median :61574   Median :39.0   Median :26.63   Median :116  
 Mean   :61623   Mean   :40.3   Mean   :27.58   Mean   :118  
 3rd Qu.:66765   3rd Qu.:56.0   3rd Qu.:31.30   3rd Qu.:127  
 Max.   :71915   Max.   :80.0   Max.   :81.25   Max.   :226

Visualization

Histogram

# Filter out non-finite values from the BMI column
nhanes_subset <- nhanes_subset %>%
  filter(is.finite(BMI))

# Create a histogram of BMI with more ticks on the axes
ggplot(nhanes_subset, aes(x = BMI)) +
  geom_histogram(binwidth = 2, fill = "red", color = "white") +
  ggtitle("Histogram of BMI") +
  xlab("BMI") +
  ylab("Frequency") +
  theme_minimal() + # Apply a minimal theme for a clean look
  theme(
    plot.background = element_rect(fill = "white"),
    panel.background = element_rect(fill = "white")
  ) +
  xlim(0, 60) + # Trim the x-axis to show values between 10 and 60
  ylim(0, 800) +  # Trim the y-axis to show values between 0 and 800
  scale_x_continuous(breaks = seq(0, 60, by = 10), limits = c(10, 60)) + # More ticks on the x-axis
  scale_y_continuous(breaks = seq(0, 800, by = 50)) # More ticks on the y-axis

Scale for x is already present.
Adding another scale for x, which will replace the existing scale.
Scale for y is already present.
Adding another scale for y, which will replace the existing scale.

Warning: Removed 12 rows containing non-finite outside the scale range
(`stat_bin()`).

Warning: Removed 2 rows containing missing values or values outside the scale range
(`geom_bar()`).

Scatter Plot:

# Create a scatter plot of BMI vs BloodPressure
# Sample 10% of the data points
nhanes_sampled_data <- nhanes_subset %>% sample_frac(0.1)

# Create a scatter plot of BMI vs BloodPressure with the sampled data
ggplot(nhanes_sampled_data, aes(x = BMI, y = BloodPressure)) +
  geom_point(shape = 19, color = "blue") +
  ggtitle("Scatter Plot of BMI vs Blood Pressure") +
  xlab("BMI") +
  ylab("Blood Pressure") +
  theme_minimal() + # Apply a minimal theme with a white background
  theme(
    plot.background = element_rect(fill = "white"),
    panel.background = element_rect(fill = "white")
  )

Box Plot:

# Create a box plot of BloodPressure by Age groups:
nhanes_sampled_data <- nhanes_sampled_data %>%
                       mutate(AgeGroup = cut(Age, breaks = c(20, 40, 60, 80), 
                                             labels = c("20-40", "40-60", "60-80")))

# Create a box plot of BloodPressure by Age groups with colors
ggplot(nhanes_sampled_data, aes(x = AgeGroup, y = BloodPressure, fill = AgeGroup)) +
  geom_boxplot() +
  ggtitle("Box Plot of Blood Pressure by Age Group") +
  xlab("Age Group") +
  ylab("Blood Pressure") +
  scale_fill_manual(values = c("20-40" = "red", "40-60" = "green", "60-80" = "blue")) +
  theme_minimal() +
  theme(
    plot.background = element_rect(fill = "white"),
    panel.background = element_rect(fill = "white")
  )

Line Plot:

# For this example, we'll simulate some time series data
time_series_data <- data.frame(
  Time = 1:100,
  Value = cumsum(rnorm(100))
)
# Create a line plot of the time series data
ggplot(time_series_data, aes(x = Time, y = Value)) +
  geom_line(color = "orange", linewidth = 1.5) +
  ggtitle("Line Plot of Simulated Time Series Data") +
  xlab("Time") +
  ylab("Value") +
  theme_minimal() +
  theme(
    plot.background = element_rect(fill = "white"),
    panel.background = element_rect(fill = "white")
  )

—————————————Assignment 1: Part 1——————————————————–

Dr. Kate is investigating an illness affecting a group of individuals. She needs your help to visualize the nhanes data and uncover hidden patterns.

Scatter Plot Investigation: Dr. Kate suspects that age might influence BMI in her patients. Create a scatter plot using the nhanes dataset to visualize the relationship between Age and BMI. Identify any patterns or trends that might suggest an age-related trend in BMI.

# Solution a


# Hint: Look for clusters or patterns suggesting that certain age groups are more prone to higher or lower BMI, indicating lifestyle or metabolic factors.

Box Plot Analysis: Dr. Kate wants to determine if certain age groups are at higher risk for elevated blood pressure. Generate a box plot for BloodPressure by AgeGroup in the nhanes dataset. Identify any age groups with higher or more variable blood pressure, indicating a health risk.

# Solution b


# Hint: Notice any age group with outliers or higher blood pressure, suggesting that this group is at higher risk of the mysterious illness.

Histogram Examination: Dr. Kate is also interested in the distribution of BMI among her patients. Create a histogram of BMI. Analyze the distribution of BMI levels and identify common ranges and any outliers.

# Solution c


# Hint: Identify common BMI ranges and any outliers. Individuals with significantly high or low BMI might be suffering from the mysterious illness, possibly linked to trends observed in blood pressure and age group