A comparison of methods for differential expression analysis of RNA-seq data

By Charlotte Soneson, Mauro Delorenzi. BMC Bioinformatics, 2013

891 citations recorded by the source corpus.

Finding genes that are differentially expressed between conditions is an integral part of understanding the molecular basis of phenotypic variation. In the past decades, DNA microarrays have been used extensively to quantify the abundance of mRNA corresponding to different genes, and more recently high-throughput sequencing of cDNA (RNA-seq) has emerged as a powerful competitor. As the cost of sequencing decreases, it is conceivable that the use of RNA-seq for differential expression analysis will increase rapidly. To exploit the possibilities and address the challenges posed by this relatively new type of data, a number of software packages have been developed especially for differential expression analysis of RNA-seq data. We conducted an extensive comparison of eleven methods for differential expression analysis of RNA-seq data. All methods are freely available within the R framework and take as input a matrix of counts, i.e. the number of reads mapping to each genomic feature of interest in each of a number of samples. We evaluate the methods based on both simulated data and real RNA-seq data. Very small sample sizes, which are still common in RNA-seq experiments, impose problems for all evaluated methods and any results obtained under such conditions should be interpreted with caution. For larger sample sizes, the methods combining a variance-stabilizing transformation with the ‘limma’ method for differential expression analysis perform well under many different conditions, as does the nonparametric SAMseq method.

Subfields: Molecular Biology

View Paper: https://www.semanticscholar.org/paper/6ba13c951fdc3cc472a8ceeba02142299236f1f4

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Molecular Biology

Citations in the corpus

Count-based differential expression analysis of RNA sequencing data using R and Bioconductor (2013, methodology): “Here R and Bioconductor are adopted to tie together the workflow and provide data structures, and their unique strengths in workflow components are leveraged, including statistical algorithms, visualization”
Sailfish enables alignment-free isoform quantification from RNA-seq reads using lightweight algorithms (2013, background): “one of the most computationally time-consuming steps, and it is commonly the first step in an analysis of differential expression among multiple samples [13].”
How many biological replicates are needed in an RNA-seq experiment and which differential expression tool should you use? (2015, background): “2013; Soneson 2014), or a combination of the two (Kvam et al. 2012; Li et al. 2012; Dillies et al. 2013; Guo et al. 2013; Soneson and Delorenzi 2013; Burden et al. 2014).”
Effects of library size variance, sparsity, and compositionality on the analysis of microbiome data (2015, background): “6a), suggesting a 541 high false positive rate in agreement with RNA-Seq studies (Love MI 2014; Rapaport et al. 542 2013; Soneson & Delorenzi 2013).”
voom: precision weights unlock linear model analysis tools for RNA-seq read counts (2014, methodology): “Some of the count-based methods have been criticized as being sensitive to outlier counts [28].”
Normalization and microbial differential abundance strategies depend upon data characteristics (2017, result): “In simulations where the abundances of 10% of the OTUs increased in one group, all but ANCOM [7] had a highly inflated average FDR, in some cases exceeding 40% (Fig.”
Comparison of software packages for detecting differential expression in RNA-seq studies (2013, methodology): “Unlike Soneson and Delorenzi [16], we also included into our systematic comparisons NOIseq and Cuffdiff 2, both of which performed surprisingly poorly in the present data sets.”
Systems biology and gene networks in neurodevelopmental and neurodegenerative disorders (2015, background): “• Appropriately transformed and normalized sequencing data can be treated similarly to microarray data as far as statistical modelling and multiple corrections are concerned 175,220 .”
SARTools: A DESeq2- and EdgeR-Based R Pipeline for Comprehensive Differential Analysis of RNA-Seq Data (2015, methodology): “In recent years edgeR and a previous version of DESeq2, DESeq [4], have been included in several benchmark studies [5, 6] and have shown to perform well in replicated experiments.”
Unifying the analysis of high-throughput sequencing datasets: characterizing RNA-seq, 16S rRNA gene sequencing and selective growth experiments by compositional data analysis (2014, methodology): “Both DESeq and baySeq were recently shown to be among the most conservative when examining this dataset [36], and so can be considered to have assumptions that have been iteratively altered to fit to the underlying data better than the majority of tools used for this purpose.”
A survey of best practices for RNA-seq data analysis (2016, methodology): “Most comparison studies have focused on simulated datasets [56, 208, 209] or on samples to which exogenous RNA ( ‘ spike-in ’ ) has been added in known quantities [63, 196].”
RNA Sequencing Data: Hitchhiker's Guide to Expression Analysis (2018, methodology): “In order to avoid too liberal inference, the dispersion is then set as the maximum between the smooth fit and the genewise MM estimate ; however, while robust to outliers, this method tends to overestimate the variance and is therefore conservative [135,136] .”
A Comparative Study of Techniques for Differential Expression Analysis on RNA-Seq Data (2014, background): “Recently, several published studies have evaluated the performance of RNA-Seq software tools for differential expression analysis, although they all confined their analyses to methods that are applicable to count matrices (also known as count tables that contain raw counts for each feature) [28-30].”
Robustly detecting differential expression in RNA sequencing data using observation weights (2013, methodology): “(6,7)) and those that accompany every new method published, is that they are a snapshot in time.”
Corset: enabling differential gene expression analysis for de novo assembled transcriptomes (2014, methodology): “Similarly, the choice of method for performing count-based statistical testing for differential expression has been evaluated [16,17].”
Data quality aware analysis of differential expression in RNA-seq with NOISeq R/Bioc package (2015, methodology): “Since NOISeqBIO is a non-parametric method, we also included a non-parametric method in the comparison: SAMseq (38), which has been reported to perform well (20).”