Differential expression analysis using edgeR

Name: Differential expression analysis using edgeR
Rating: 4.4 (62 reviews)

Learn how to compare two biological conditions using transcriptomic data.

Created byGeorge Sentis

Last updated 10/2023

English

English [Auto],

What you'll learn

Understand the structure of RNA-Seq data objects in R
Explore your RNA-Seq data
Clean your RNA-Seq data
Perform differential expression analysis using edgeR

Course content

5 sections • 15 lectures • 1h 52m total length

Introduction3:11
beginner-friendly introduction to differential expression analysis using edgeR in R, guiding RNA-seq data exploration with read counts, sample information, and a psoriatic arthritis dataset from the Gene Expression Omnibus repository.
R studio1:45
Install the language and use RStudio, an integrated development environment favored by bioinformaticians, then explore the script, environment, plots, console, installed packages, and help on functions.
Refreshing basic R commands2:32
Master R commands: check and change your working directory with getwd and setwd, switch to home with tilde, assign and remove variables with <-, =, and rm to clear workspace.
Installing the packages2:25
Install the required packages from cran and bioconductor using the bike manager, then install the remaining packages to start the differential expression analysis with edgeR.

Let's start! Loading packages and files6:55
Load and initialize essential libraries, including open zsh for excel manipulation, ggplot, dplyr, factominer, and biomart. Read expression read counts and sample information to prepare data frames for downstream analysis.
Data preparation5:27
Reorder expression data to match sample metadata, verify with identical, and check for missing values using is.na before proceeding to normalize counts in edgeR differential expression analysis.
Model matrix and expression check10:55
Build the edgeR model matrix by treating tissue type as a factor, compare healthy, lesional, and uninvolved skin, and verify normalization with CPM and log CPM.
Count filtering and normalization7:28
Learn to filter lowly expressed genes in edgeR by keeping genes with CPM > 1 in at least 5% of samples, then re-normalize and compute log counts per million.
Exploratory data analysis16:10
Explore data visually through box plots and log normalized expressions, perform a principal component analysis using FactoMineR, and visualize pca with ggplot to distinguish lesional, healthy, and uninvolved samples.

Differential expression analysis18:32
Estimate dispersions and fit a differential expression model with edgeR, define contrasts lesional versus healthy, compute log fold changes and FDR, and filter by abs logFC and FDR.
Annotation of results9:33
Annotate results with Biomart using the useEnsembl function to fetch Ensembl IDs, descriptions, positions, and symbols, then left-join annotations to your data for lesional skin versus healthy samples.

Results Visualization (Part 1): Volcano Plot10:21
Visualize differential expression with a volcano plot, plotting log fold change on the x-axis and -log10 FDR on the y-axis to identify significant upregulated and downregulated genes via thresholds.
Results Visualization (Part 2): Heatmap11:16
Plot a top 100 differential expressed gene heatmap (50 up, 50 down) using log normalized counts to show expression across lesional skin, uninvolved, and healthy controls.
Ensuring reproducibility4:51
Document edgeR differential expression analysis reproducibly by saving the complete workspace including volcano plots and heat maps with date-stamped names, recording package versions via sessionInfo, and logging output with sink.

Requirements

Basic R programming skills

Description

Hello everyone!
My name is George and I am a bioinformatician.
I am here to guide you through an analysis of RNA sequencing data using edgeR.

This is a differential expression analysis where we compare samples from Psoriatic Arthritis patients to Healthy controls in order to identify deregulated genes. It is a very simple and common experimental design that is an industry-standard nowadays and a great tool to have at your disposal.

For this course, you need to have some basic R programming experience and you need to have a computer with R language installed. This course is intended for people aiming to dive into the world of RNA sequencing data analysis and have not had the chance to analyse data using edgeR before.

We will use the publicly available Psoriatic Arthritis dataset and perform some data cleaning, preparation, exploratory analysis, data normalization and subsequently reach our desired goal which is the differential expression analysis. Finally, we will visualize the results by creating a volcano plot and a heatmap of the top differentially expressed genes and we will save any information needed for this analysis to be documented in a reproducible way.

If you fulfil the above-mentioned criteria, tag along and let's explore the basics of differential expression using edgeR!

Who this course is for:

Beginner or Advanced R users aiming to delve into bioinformatics

Differential expression analysis using edgeR

What you'll learn

Explore related topics

Course content

Introduction4 lectures • 10min

Data Prep5 lectures • 47min

Differential expression analysis2 lectures • 28min

Visualization of Results & Reproducibility3 lectures • 26min

Closing1 lecture • 1min

Requirements

Description

Who this course is for: