A scaling normalization method for differential expression analysis of RNA-seq data
By Mark D. Robinson, Alicia Oshlack. Genome Biology, 2010
6,941 citations recorded by the source corpus.
The fine detail provided by sequencing-based transcriptome surveys suggests that RNA-seq is likely to become the platform of choice for interrogating steady state RNA. In order to discover biologically important changes in expression, we show that normalization continues to be an essential step in the analysis. We outline a simple and effective method for performing normalization and show dramatically improved results for inferring differential expression in simulated and publicly available data sets.
Subfields: Cancer Research, Molecular Biology
View Paper: https://www.semanticscholar.org/paper/8bc2fc891291579dcc0da5ace37a5a9b6057e086
How researchers see this paper
The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.
In Cancer Research
- Gordon K. Smyth
- Aaron T. L. Lun
- John C. Marioni
- Yunshun Chen
- Fabian J Theis
- Michael I. Love
- Wei Shi
- Sarah A. Teichmann
In Molecular Biology
- Gordon K. Smyth
- Aaron T. L. Lun
- John C. Marioni
- Yunshun Chen
- Mark D. Robinson
- Jeffrey T. Leek
- Charlotte Soneson
- Fabian J Theis
Citations in the corpus
- Human Tumor-Associated Macrophage and Monocyte Transcriptional Landscapes Reveal Cancer-Specific Reprogramming, Biomarkers, and Therapeutic Targets (2019, methodology): “Gene expression levels were normalized using the Trimmed Mean of M-values (TMM) method using the calcNormFactors() function and log2 transformed using the cpm() function from the EdgeR package in R (Robinson and Oshlack, 2010).”
- RNA-Seq Signatures Normalized by mRNA Abundance Allow Absolute Deconvolution of Human Immune Cell Types (2019, methodology): “values that optimize the error between deconvolution results and flow cytometry proportions (Figure 7C; STAR Methods and Absolute Deconvolution), inverted HKmean values (Figure S7D), and inverted TMM values (RNATMM) (Robinson and Oshlack, 2010) (Figure S7E; STAR Methods).”
- Pooling across cells to normalize single-cell RNA sequencing data with many zero counts (2016, methodology): “This is the underlying concept of commonly used methods such as DESeq [3] and trimmed mean of M values (TMM) normalization [4].”
- Differential expression analysis of multifactor RNA-Seq experiments with respect to biological variation (2012, background): “The library size may be defined as the total number of mapped reads, or it may be estimated from the data to effect some relative normalization between the different libraries (26).”
- voom: precision weights unlock linear model analysis tools for RNA-seq read counts (2014, methodology): “TMM-scale normalization [54] was used for all the analysis methods, except for DESeq and PoissonSeq, which have their own built-in normalization methods.”
- From reads to genes to pathways: differential expression analysis of RNA-Seq experiments using Rsubread and the edgeR quasi-likelihood pipeline (2016, methodology): “Normalization by trimmed mean of M values (TMM)17 is performed by using the calcNormFactors function, which returns the DGEList argument with only the norm.factors changed.”
- GC-Content Normalization for RNA-Seq Data (2011, background): “Some authors have argued that it is better to leave the count data unchanged to preserve their sampling properties and instead use an offset for normalization purposes in the statistical model for read counts [21,22].”
- Gene co-expression analysis for functional classification and gene–disease predictions (2017, methodology): “TMM has been suggested as a better alternative [58].”
- Scater: pre-processing, quality control, normalization and visualization of single-cell RNA-seq data in R (2017, methodology): “Frequently used methods for scaling normalization include the trimmed mean of Mvalues (Robinson and Oshlack, 2010), relative log-expression (Anders and Huber, 2010) and upper-quartile methods (Bullard et al.”
- Selecting between‐sample RNA‐Seq normalization methods from the perspective of their assumptions (2016, methodology): “TMM sets one of the samples as a reference sample, then compares the counts in each sample to the reference sample to estimate the ratio of sequencing depths between each sample and the reference.”
- Data quality aware analysis of differential expression in RNA-seq with NOISeq R/Bioc package (2015, methodology): “Technology biases, such as the transcript length (12), GC content (13), PCR artifacts, uneven transcript read coverage, contamination by off-target transcripts, or large differences in transcript distributions (14), are factors that interfere in the linear relationship between transcript abundance a”
- An atlas of the protein-coding genes in the human, pig, and mouse brain (2020, methodology): “The within-sample normalization was based on protein-coding transcripts per million (pTPM), while the between-sample normalization was based on trimmedmeans ofM values (TMM) (22), Pareto scaling per gene (23), and limma (24), resulting in normalized expression (NX) values calculated for all genes ac”
- Differential abundance testing on single-cell data using K-nearest neighbour graphs (2021, methodology): “To incorporate complex experimental designs (for example, the presence of multiple conditions), we test for differences in abundance using a negative binomial (NB) GLM framework 10,11 , normalizing for differences in cell numbers across samples 12 (Methods and Supplementary Notes).”
- Normalizing single-cell RNA sequencing data: Challenges and opportunities (2017, methodology): “Two highly popular methods are DESeq [22] and Trimmed Mean of M-values (TMM) [23] normalization.”
- From RNA-seq reads to differential expression results (2010, methodology): “So far, quantile normalization does not seem to improve DE detection to the same extent as an appropriate scaling factor [45] and it is not clear that the power law distribution applies to all datasets [48].”