Lesson objectives

Understand the dataset being used
Setup an R project, import the data files and do a first exploration of what they are

Summary of dataset

In this lesson, we will apply some of the skills that we’ve gained so far to manipulate and explore a dataset from an RNAseq experiment.

This lesson uses data from an experiment included in the fission R/Bioconductor package. Very briefly, we have transcriptome data for:

Two yeast strains: wild type (“wt”) and atf21del mutant (“mut”)
Each has 6 time points of osmotic stress time (0, 15, 30, 60, 120 and 180 mins)
Three replicates for each strain at each time point

Let’s say that you did this experiment yourself, and that a bioinformatician analysed it and provided you with four files of data:

sample_info.csv - information about each sample.
counts_raw.csv - “raw” read counts for all genes, which gives a measure of the genes’ expression. (these are simply scaled to the size of each library to account for the fact that different samples have more or less total number of reads).
counts_transformed.csv - normalised read counts for all genes, on a log scale and transformed to correct for a dependency between the mean and the variance. This is typical of count data, and we will look at it in the exploratory data analysis lesson).
test_result.csv - results from a statistical test that assessed the probability of observed expression differences
between the first and each of the other time points in WT cells, assuming a null hypothesis of no difference.

Getting started

The data are provided as CSV files, which you can download and read into your R session.

create a new RStudio project in a new directory (File > New Project...).
create a new folder called scripts
create a new script called 01_prepare_data.R.

Use the code below to download the data into a new directory called data:

# Create a "data" directory
dir.create("data")

# Download the data provided by your collaborator
# using a for loop to automate this step
for(i in c("counts_raw.csv", "counts_transformed.csv", "sample_info.csv", "test_result.csv")){
  download.file(
    url = paste0("https://github.com/tavareshugo/data-carpentry-rnaseq/blob/master/data/", i, "?raw=true"),
    destfile = paste0("data/", i)
  )
}

Finally, load the tidyverse package and do the exercises below:

# load the package
library(tidyverse)

Exercise:

Import data into R and familiarise yourself with it.

Create four objects called raw_cts, trans_cts, sample_info and test_result.

Link to full exercise

back to lesson’s homepage

Exploratory analysis of RNAseq data

Lesson objectives

Summary of dataset

Getting started