Accurate differential gene expression analysis for RNA-Seq data without replicates

Despite the fact that many RNA-Seq experiments do not use biological replicates, most of the existing differential gene expression analysis for RNA-Seq data are designed to work with replicates. We present a novel method for differential gene expression analysis based on bootstrapping. We also discuss the use of different normalization methods with Fisher’s exact test. We compare the methods we present, with existing methods, on real RNA-Seq benchmarking data. Our comparison shows that bootstrapping outperforms other methods in most cases.