Exercise 2 — Introduction to Data Frames and Sample Metadata

In real bioinformatics workflows (especially RNA-seq), data is rarely stored as simple vectors.

Instead, we use data frames, which allow us to store:

  • gene expression values
  • sample IDs
  • conditions (Control vs Treatment)
  • batch information
  • clinical metadata

This exercise introduces the most important structure in R: the data frame.

Part 1 — Creating a Data Frame

We will simulate gene expression data from 8 samples.

Task 1.1

Create the following vectors:

gene <- c("TP53")
sample_id <- c("S1","S2","S3","S4","S5","S6","S7","S8")
expression <- c(12.4, 15.1, 8.7, 20.3, 17.8, 5.2, 13.5, 18.1)
condition <- c("Control","Control","Control","Treatment","Treatment","Treatment","Control","Treatment")

Task 1.2

Combine them into a data frame called df.

Part 2 — Inspecting the Data

Task 2.1

View the dataset.

Task 2.2

Use the following functions:

  • str(df)
  • head(df)
  • dim(df)
  • summary(df)

Questions:

  1. How many rows and columns are present?
  2. What type of data is stored in each column?

Part 3 — Accessing Columns

Task 3.1

Extract the expression column using:

  • $ operator
  • bracket notation

Task 3.2

Extract the condition column.

Task 3.3

Print only sample_id and expression.

Part 4 — Filtering Data

Filtering is extremely important in RNA-seq analysis.

Task 4.1

Find all samples where expression > 15.

Task 4.2

Find all samples in the Treatment group.

Task 4.3

Find samples where:

  • condition = “Control” AND
  • expression < 10

Task 4.4

Count:

  • number of Control samples
  • number of Treatment samples

Part 5 — Adding New Columns

Task 5.1

Create a new column called high_expression:

  • TRUE if expression > 15
  • FALSE otherwise

Task 5.2

Add another column called log_expression:

log2(expression)

Task 5.3

Inspect the updated data frame.

Part 6 — Summarising Data

Task 6.1

Calculate:

  • mean expression in Control group
  • mean expression in Treatment group

Task 6.2

Which group has higher expression?

Task 6.3

Find:

  • maximum expression per group
  • minimum expression per group

Part 7 — Biological Interpretation

Answer in comments:

# 1. Do Treatment samples generally show higher expression?
# 2. What does "high expression" mean biologically here?
# 3. Why do we use log transformation in RNA-seq?

Bonus Challenge

Task B1

Sort the data frame by expression (highest to lowest).

Task B2

Create a subset containing only Treatment samples.

Task B3

Add a new column:

expression_scaled = expression / max(expression)
⬅️ Previous Exercise 🏠 Home ➡️ Next Exercise