DE analysis of RNA-seq data

For RNA-seq data, the strategy taken is to count the number of reads that fall into annotated genes and to perform statistical analysis on the table of counts to discover quantitative changes in expression levels between experimental groups. Easy, right? Not, exactly.

1. We have integer counts and not continous measurements. Data is not normally distributed, so statistical methods we applied to microarray data don't work here.

2. Replication levels in designed RNA-Seq experiments tend to be modest, often not much more than two or three. As a result, there is a need for statistical methods that perform well in small-sample situations.

3. There is a dependence of variance on the mean (which changes with increasing number of replicates)

Solution: Appropriate modelling of the mean-variance relationship in DGE data is important for making inferences about differential expression. Employing methods which assess the mean-variance relationship to help with the problem of estimating biological variability for experiments with small number of replicates.

In this module, learners will use R Statistical Software to walk-through activities designed to compare the performance of different tools (edgeR, DESeq2, limma-voom) for differential expression analysis of RNA-Seq data, and how the mean-variance relationship is addressed in datasets with increasing number of replicates.

Learning Objectives

• understanding the relationship between mean and variance and how that changes with the number of replicates
• familiarizing yourself with tools for DE analysis in R
• understanding the importance of having replicates in a RNA-Seq study