Moderated statistical tests for assessing differences in tag abundance

By Mark D. Robinson, Gordon K. Smyth. Bioinform., 2007

924 citations recorded by the source corpus.

MOTIVATION Digital gene expression (DGE) technologies measure gene expression by counting sequence tags. They are sensitive technologies for measuring gene expression on a genomic scale, without the need for prior knowledge of the genome sequence. As the cost of sequencing DNA decreases, the number of DGE datasets is expected to grow dramatically. Various tests of differential expression have been proposed for replicated DGE data using binomial, Poisson, negative binomial or pseudo-likelihood (PL) models for the counts, but none of the these are usable when the number of replicates is very small. RESULTS We develop tests using the negative binomial distribution to model overdispersion relative to the Poisson, and use conditional weighted likelihood to moderate the level of overdispersion across genes. Not only is our strategy applicable even with the smallest number of libraries, but it also proves to be more powerful than previous strategies when more libraries are available. The methodology is equally applicable to other counting technologies, such as proteomic spectral counts. AVAILABILITY An R package can be accessed from http://bioinf.wehi.edu.au/resources/

Subfields: Language and Linguistics

View Paper: https://www.semanticscholar.org/paper/e70ba61fab6dfbc4506fc86ed19698cfd9db9384

How researchers see this paper

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

In Language and Linguistics

Citations in the corpus

A comparison of statistical methods for detecting differentially expressed genes from RNA-seq data. (2012, positive): “[21] To estimate the dispersion in either case edgeR implements a quantile-adjusted conditional maximum likelihood (qCML) method.”
A Comparative Study of Techniques for Differential Expression Analysis on RNA-Seq Data (2014, background): “However, when biological replicates are included, it has been noted that the Poisson distribution underestimates the variation seen in the data [20,21], a problem known as overdispersion.”
Finding consistent patterns: A nonparametric approach for identifying differential expression in RNA-Seq data (2013, methodology): “These methods are parametric: they assume that (maybe after some simple transformation) each Nij is drawn from a particular distribution, such as Gaussian ([4, 12]), Poisson ([16, 7, 32, 11, 14]), negative binomial ([23, 24, 22, 11, 1]), et al ([2, 15]).”
DiBind : Dierential binding analysis of ChIP-Seq peak data (2016, methodology): “The final steps are to perform testing to determine the significance measure of the differences between the sample groups by calling exactTest ([6]) using the DGEList with the dispersion set based on the bTagwise parameter.”
Differential analysis of RNA-seq incorporating quantification uncertainty (2016, methodology): “Figure 2 shows the result of our method in identifying differentially expressed genes, alongside results from Cuffdiff 2 24 , DESeq 3 , DESeq2 19 , EBSeq 15 , edgeR 11,18 , voom 5 , and log fold change 20 .”
A comparison of methods for differential expression analysis of RNA-seq data (2013, methodology): “A common dispersion for all genes may however be a too restrictive assumption, and so this procedure was developed further to allow for gene-wise dispersion estimates, but where the individual estimates were squeezed towards the common one using a weighted likelihood approach [31].”
Evaluation of statistical methods for normalization and differential expression in mRNA-Seq experiments (2010, methodology): “The use of negative binomial or empirical Bayes methods, as described in the SAGE literature [21,22], may be sensible in such settings of increased variability.”
Moderated estimation of fold change and dispersion for RNA-seq data with DESeq2 (2014, positive): “edgeR [2, 3] moderates the dispersion estimate for each gene toward a common estimate across all genes, or toward a local estimate from genes with similar expression strength, using a weighted conditional likelihood.”
Differential expression analysis for sequence count data (2010, methodology): “…we discovered that the use of total read counts as estimates of sequencing depth, and hence for the adjustment of observed counts between samples (as recommended by Robinson et al. [8] and others) may result in For DESeq , no excess of low P values was seen, as expected when comparing replicates.”
Count-based differential expression analysis of RNA sequencing data using R and Bioconductor (2013, methodology): “Simon Anders 1 , Davis J McCarthy 2,3 , Yunshun Chen 4,5 , Michal Okoniewski 6 , Gordon K Smyth 4,7 , Wolfgang Huber 1 & Mark D Robinson 8,9 make formal distributional assumptions about the data observed.”
RNA-seq: an assessment of technical reproducibility and comparison with gene expression arrays. (2008, background): “www.genome.org bution, such as the quasi-Poisson distribution (Venables and Ripley 2002) or the negative binomial distribution (Robinson and Smyth 2007), which have an additional parameter that estimates over-(or under-) dispersion relative to a Poisson model.”
Detecting differential usage of exons from RNA-seq data (2012, methodology): “Subsequently, Robinson and coworkers presented the edgeR method (Robinson and Smyth 2007, 2008; Robinson et al. 2010b), which introduced the use of the negative binomial distribution to RNA-seq analysis.”
From reads to genes to pathways: differential expression analysis of RNA-Seq experiments using Rsubread and the edgeR quasi-likelihood pipeline (2016, methodology): “This method provides stricter error rate control than other negative binomial based pipelines, including the traditional edgeR pipelines [8, 9, 6] or DESeq2 [10].”
baySeq: Empirical Bayesian methods for identifying differential expression in sequence count data (2010, methodology): “Since in SAGE data, there is only one tag per mRNA molecule, Robinson and Smyth [9] examine methods for detecting differentially expressed tags between samples.”