Skip to content

Differential Expression Analysis

Differential expression analysis means taking RNA-Seq expression data and performing statistical analysis to discover quantitative changes in gene expression levels between experimental groups. In other words, differential expression analysis consists of the identification of genes (or other types of genomic features, such as transcripts or exons) that are expressed in significantly different quantities in distinct groups of samples. There are several statistical methodologies that allow addressing different experimental designs: biological conditions, diseased vs healthy, different tissues, different development stages, different gender, time series…

Four strategies are available:

  • Pairwise Differential Expression Analysis: Pairwise differential expression analysis allows the identification of differentially expressed genes considering different experimental conditions studied in RNA-Seq experiments. It is based on the software package "edgeR" (empirical analysis of DGE in R), which belongs to the Bioconductor project.
  • Differential Expression Analysis with DESeq2: Based on the "DESeq2" Bioconductor package, this application models RNA-Seq count data using a negative binomial distribution, estimating gene-wise dispersion and fold changes to test for differential expression between experimental conditions.
  • Differential Expression Analysis with limma: Based on the "limma" Bioconductor package, this application fits a linear model to each feature and applies empirical Bayes moderation of the variance. RNA-seq counts are first transformed with voom, allowing the same statistical framework used for microarrays to also be applied to RNA-Seq experiments.
  • Pairwise Differential Expression Analysis (Without Replicates): Pairwise differential expression analysis (without replicates) is suited when there are not replicates for any of the experimental conditions. It is based on the Bioconductor software package "NOISeq", which can compare samples from two experimental conditions by simulating replicates.