The DOI matches its Crossref record; 22 of the 23 references it cites were found in a registry or an index.
Evaluation of statistical methods for normalization and differential expression in mRNA-Seq experiments
Open accessAbstract
BACKGROUND: High-throughput sequencing technologies, such as the Illumina Genome Analyzer, are powerful new tools for investigating a wide range of biological and medical questions. Statistical and computational methods are key for drawing meaningful and accurate conclusions from the massive and complex datasets generated by the sequencers. We provide a detailed evaluation of statistical methods for normalization and differential expression (DE) analysis of Illumina transcriptome sequencing (mRNA-Seq) data. RESULTS: We compare statistical methods for detecting genes that are significantly DE between two types of biological samples and find that there are substantial differences in how the test statistics handle low-count genes. We evaluate how DE results are affected by features of the sequencing platform, such as, varying gene lengths, base-calling calibration method (with and without phi X control lane), and flow-cell/library preparation effects. We investigate the impact of the read count normalization method on DE results and show that the standard approach of scaling by total lane counts (e.g., RPKM) can bias estimates of DE. We propose more general quantile-based normalization procedures and demonstrate an improvement in DE detection. CONCLUSIONS: Our results have significant practical and methodological implications for the design and analysis of mRNA-Seq experiments. They highlight the importance of appropriate statistical methods for normalization and DE inference, to account for features of the sequencing platform that could impact the accuracy of results. They also reveal the need for further research in the development of statistical and computational methods for mRNA-Seq.
Overview
The claims and statistics of this paper haven’t been checked yet.
Run the check — free, no account needed. It reads the open-access text and checks the sources it cites.
Integrity signals
Registry
Cites
Statistics
Figures
![Figure 4 Comparison of biological, library preparation, and flow-cell effects. Boxplots of estimated log-fold-changes for UHR vs. Brain biological effects (GLM 2 in [Additional file 2: Supplemental Table S4]), flow-cell effects adjusting for biology (GLM 4), library preparation effects within flow-cell (GLM 7). Estimates are presented for total-count (black) and upper-quartile (blue) normalization.](/paper-figures/7a218023bc1142f6aea616fd271add05/evaluation-of-statistical-methods-for-normalization-and-differential-expression-figure-4.jpg)
![Figure 6 Comparison of mRNA-Seq and microarray differential expression calls to qRT-PCR: ROC curves. Genes common to all three platforms and present for both qRT-PCR and sequencing (see [Additional file 2: Supplemental Section S6]). evaluated and declared DE if their qRT-PCR absolute log-ratio was (a) greater than 2 or (b) greater than 0.5; genes were declared non-DE if their absolute log-ratio was less than 0.2. The GLM-based likelihood ratio test was used for the sequencing data. Two normalization procedures are presented for mRNA-Seq: total-count (black) and upper-quartile (blue) normalization. Microarray data were normalized using RMA (gray). Note that we require a true positive to be differentially expressed in the same direction according to both mRNA-Seq and qRT-PCR (see Table 1 and Methods).](/paper-figures/e79890d44a604905bdf241f0104817db/evaluation-of-statistical-methods-for-normalization-and-differential-expression-figure-6.jpg)
![Figure 7 Comparison of mRNA-Seq and microarray differential expression measures to qRT-PCR. Difference scatterplots comparing the estimates of UHR/Brain expression log-ratio from qRT-PCR to those from (a) mRNA-Seq, using the standard total-count normalization, and (b) microarrays, using the standard RMA normalization. Shown are the genes shared between all three platforms, present in both Brain and UHR according to both mRNA-Seq and qRT-PCR (see [Additional file 2: Supplemental Section S1]), and having absolute qRT-PCR expression log-ratio less than 4. Horizontal lines in (a) represent the median UHR/Brain log-ratio for the sequencing data after the standard total-count normalization (black), POLR2A normalization (red), quantile normalization (yellow), upper-quartile normalization (blue); horizontal lines in (b) show the median UHR/Brain log-ratio for the microarray data after the standard RMA normalization (black) and POLR2A normalization (red).](/paper-figures/ffe14db2d64a46939165f0c6e8ca7bec/evaluation-of-statistical-methods-for-normalization-and-differential-expression-figure-7.jpg)
![Figure 8 Comparison of normalization procedures: Goodness-of-fit of Poisson model. The multiplicative Poisson model (GLM 1 in [Additional file 2: Supplemental Table S4]) is fit to the seven Brain lanes in the MAQC-2 experiment after (a) total-count, (b) POLR2A, (c) upper-quartile, and (d) quantile normalization. Goodness-of-fit statistics are computed and displayed in χ2 quantile-quantile plots. Genes with goodness-of-fit statistics in the top quantiles of the χ2-distribution are displayed using colored plotting symbols: red (1, 5]%, purple (.1, 1]%, gold [0, .1]%. Similar plots for UHR show the same patterns.](/paper-figures/d99eb5ce90ee4a9ebac8c9f438b4c283/evaluation-of-statistical-methods-for-normalization-and-differential-expression-figure-8.jpg)
Verification
Claims the paper makes
The claims and statistics of this paper haven’t been checked yet.
Run the check — free, no account needed. It reads the open-access text and checks the sources it cites.
Claims other papers make about this one
Each is checked against this paper’s own text: does it say what the citing paper says it does?
The source backs part of the claim, not all of it — or backs it for a different population, condition or comparison.
In the citing paper
“These approaches rely on normalizing methods that are based on total or effective counts, and tend to perform poorly when samples have heterogeneous transcript distributions, that is, when highly and differentially expressed features can skew the count distribution [45, 46].”
From A survey of best practices for RNA-seq data analysis (2016) · cites this paper as [45]
Experimental — accuracy not yet measuredIn the citing paper
“While reproducibility in RNA-seq is usually high at the level of sequencing [1, 45], other steps such as RNA extraction and library preparation are noisier and may introduce biases in the data that can be minimized by adopting good experimental procedures (Box 2).”
From A survey of best practices for RNA-seq data analysis (2016) · cites this paper as [45]
Self-check
Percentages, sums and reported differences re-computed from the paper’s own numbers.
The self-check needs the paper’s full text.
Citations
Cites
Ultrafast and memory-efficient alignment of short DNA sequences to the human genome.
Langmead, B., Trapnell, C., Pop, M., & Salzberg, S. L. (2009) · Genome Biology · doi:10.1186/gb-2009-10-3-r25 · Cited by 21,849 · cited 1×
“We used Bowtie to map reads to the genome (GRCh37 assembly) [ 16 ].”
Method(2005) · Springer eBooks · doi:10.1007/0-387-27255-0_9 · Cited by 20,333
Mapping and quantifying mammalian transcriptomes by RNA-Seq
Mortazavi, A., Williams, B. A., McCue, K., Schaeffer, L., & Wold, B. (2008) · Nature Methods · doi:10.1038/nmeth.1226 · Cited by 13,696 · cited 6×
“These high-throughput sequencing technologies have already been applied to monitor genome-wide transcription levels (mRNA-Seq), DNA-protein interactions (ChIP-Seq), chromatin structure, and DNA methylation [1-9].”
BackgroundMethodInfluentialSmyth, G. K. (2004) · Statistical Applications in Genetics and Molecular Biology · doi:10.2202/1544-6115.1027 · Cited by 11,829
Exploration, normalization, and summaries of high density oligonucleotide array probe level data
Irizarry, R. A. (2003) · Biostatistics · doi:10.1093/biostatistics/4.2.249 · Cited by 10,603
Base-Calling of Automated Sequencer Traces Using Phred. I. Accuracy AssessmentPreprint
Ewing, B., Hillier, L., Wendl, M. C., & Green, P. (1998) · Genome Research · doi:10.1101/gr.8.3.175 · Cited by 6,993
Base-Calling of Automated Sequencer Traces Using Phred. II. Error Probabilities
Ewing, B., & Green, P. (1998) · Genome Research · doi:10.1101/gr.8.3.186 · Cited by 5,439 · cited 1×
“quality scores for the other seven lanes [[15], Supplementary Information, p. 7]. Bustard can also be run using the auto-calibration method, which scores basecalls in a manner similar to the phred base-caller [ 17 ] and does not require a control lane per flow-cell.”
MethodAlternative isoform regulation in human tissue transcriptomes
Wang, E. T., Sandberg, R., Luo, S., Khrebtukova, I., Zhang, L., Mayr, C., Kingsmore, S. F., Schroth, G. P., & Burge, C. B. (2008) · Nature · doi:10.1038/nature07509 · Cited by 5,181
Accurate whole human genome sequencing using reversible terminator chemistry
Bentley, D. R., Balasubramanian, S., Swerdlow, H. P., Smith, G. P., Milton, J., Brown, C. G., Hall, K. P., Evers, D. J., Barnes, C. L., Bignell, H. R., et al. (2008) · Nature · doi:10.1038/nature07517 · Cited by 3,684 · cited 1×
“quality scores for the other seven lanes [[15], Supplementary Information, p.”
BackgroundRNA-seq: An assessment of technical reproducibility and comparison with gene expression arrays
Marioni, J. C., Mason, C. E., Mane, S. M., Stephens, M., & Gilad, Y. (2008) · Genome Research · doi:10.1101/gr.079558.108 · Cited by 2,771 · cited 7×
“These high-throughput sequencing technologies have already been applied to monitor genome-wide transcription levels (mRNA-Seq), DNA-protein interactions (ChIP-Seq), chromatin structure, and DNA methylation [1-9].”
BackgroundMethodResultInfluentialThe Transcriptional Landscape of the Yeast Genome Defined by RNA Sequencing
Nagalakshmi, U., Wang, Z., Waern, K., Shou, C., Raha, D., Gerstein, M., & Snyder, M. (2008) · Science · doi:10.1126/science.1158441 · Cited by 2,479 · cited 1×
“These high-throughput sequencing technologies have already been applied to monitor genome-wide transcription levels (mRNA-Seq), DNA-protein interactions (ChIP-Seq), chromatin structure, and DNA methylation [1-9].”
MethodMAQC Consortium, Shi, L., Shi, L., Reid, L. H., Jones, W. D., Shippy, R., Warrington, J. A., Baker, S. C., Collins, P. J., de Longueville, F., et al. (2006) · Nature Biotechnology · doi:10.1038/nbt1239 · Cited by 2,126
Substantial biases in ultra-short read data sets from high-throughput DNA sequencing
Dohm, J. C., Lottaz, C., Borodina, T., & Himmelbauer, H. (2008) · Nucleic Acids Research · doi:10.1093/nar/gkn425 · Cited by 1,088 · cited 2×
“0 arrays (GSE5350, samples AFX_1_ [A-B] [1-5]; see [Additional file 2: Supplemental Sections S1.”
Moderated statistical tests for assessing differences in tag abundance
Robinson, M. D., & Smyth, G. K. (2007) · Bioinformatics · doi:10.1093/bioinformatics/btm453 · Cited by 856 · cited 1×
“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.”
Method't Hoen, P. A. C., Ariyurek, Y., Thygesen, H. H., Vreugdenhil, E., Vossen, R. H. A. M., de Menezes, R. X., Boer, J. M., van Ommen, G.-J. B., & den Dunnen, J. T. (2008) · Nucleic Acids Research · doi:10.1093/nar/gkn705 · Cited by 742 · cited 2×
“literature include t-statistics with square root-transformed standard errors and Bayesian statistics based on the Beta-Binomial distribution (Hoen et al., 2008).”
MethodEvaluation of DNA microarray results with quantitative gene expression platforms
Canales, R. D., Luo, Y., Willey, J. C., Austermiller, B., Bárbácioru, C., Boysen, C., Hunkapiller, K., Jensen, R. V., Knight, C. R., Lee, K. Y., et al. (2006) · Nature Biotechnology · doi:10.1038/nbt1236 · Cited by 622 · cited 5×
“Here, we evaluate a variety of normalization procedures and focus on two main questions: (1) Does the normalization improve DE detection (sensitivity)? (2) Does the normalization result in low technical variability across replicates (specificity)? To assess DE detection, we rely on the qRT-PCR data of [13] as a gold-s…”
BackgroundMethodInfluentialHigh-resolution mapping of copy-number alterations with massively parallel sequencing
Chiang, D. Y., Getz, G. A., Jaffe, D. B., O'Kelly, M., Zhao, X., Carter, S. L., Russ, C., Nusbaum, C., Meyerson, M. M., & Lander, E. S. (2008) · Nature Methods · doi:10.1038/nmeth.1276 · Cited by 516 · cited 2×
“These high-throughput sequencing technologies have already been applied to monitor genome-wide transcription levels (mRNA-Seq), DNA-protein interactions (ChIP-Seq), chromatin structure, and DNA methylation [ 1-9 ].”
MethodInfluentialTranscript length bias in RNA-seq data confounds systems biology
Oshlack, A., & Wakefield, M. J. (2009) · Biology Direct · doi:10.1186/1745-6150-4-14 · Cited by 497 · cited 3×
“Under the assumptions presented in [11], with the unweighted t-statistics and using the same cutoff across genes, power increases with gene length for a given level of DE.”
BackgroundMethodInfluentialLi, H., Lovci, M., Kwon, Y., Rosenfeld, M. G., Fu, X., & Yeo, G. W.-M. (2008) · Proceedings of the National Academy of Sciences · doi:10.1073/pnas.0807121105 · Cited by 94 · cited 2×
“0 arrays (GSE5350, samples AFX_1_ [A-B] [1-5]; see [Additional file 2: Supplemental Sections S1.”
BackgroundMethodGenomeGraphs: integrated genomic data visualization with R
Durinck, S., Bullard, J. H., Spellman, P. T., & Dudoit, S. (2009) · BMC Bioinformatics · doi:10.1186/1471-2105-10-2 · Cited by 83 · cited 4×
“Software In order to facilitate analysis and visualization of mRNA-Seq data, we developed two R/Bioconductor software packages, Genominator and GenomeGraphs [20].”
MethodInfluentialSelf-citationLu, J., Tomfohr, J. K., & Kepler, T. B. (2005) · BMC Bioinformatics · doi:10.1186/1471-2105-6-165 · Cited by 79
Lee, A., Hansen, K. D., Bullard, J. H., Dudoit, S., & Sherlock, G. J. (2008) · PLoS Genetics · doi:10.1371/journal.pgen.1000299 · Cited by 34 · cited 3×
“The Poisson distribution has been shown to provide a good fit to the distribution of gene-level counts across replicate lanes, after normalization by total lane counts [4,6]; our experience with both the MAQC data and”
Self-citation
Research works printed but not linked: 4
- Illumina:Sequencing Analysis Software User Guide For Pipeline Version 1.3 and CASAVA Version 1.0 T. Illumina, Inc.; 2008. [Part # 1005359 Rev. A] [ http://icom.illumina.com/icom/software.ilmn?id=277 ] [Part # 1005359 Rev. A]
- Illumina:Preparing Samples for Sequencing mRNA. Ilumina, Inc.; 2009. [Part # 1004898 Rev. A] [ http://icom.illumina.com/icom/software.ilmn?id=277 ] [Part # 1004898 Rev. A]
- Taub MA: Analysis of high-throughput biological data: some statistical problems in RNA-seq and mouse genotyping. PhD thesis. Department of Statistics, UC Berkeley; 2009.
- Irizarry RA, Hobbs B, Collin F, Beazer-Barclay YD, Antonellis KJ, Scherf U, Speed TP: Exploration, normalization, and summaries of high density oligonucleotide array probe level data. Biostatistics 2003, 4(1465–4644 (Print)):249–64. 10.1093/biostatistics/4.2.249 · DOI · Paper page
Cited by
Sorted by citations
RSEM: accurate transcript quantification from RNA-Seq data with or without a reference genome
Li, B., & Dewey, C. N. (2011) · BMC Bioinformatics · doi:10.1186/1471-2105-12-323 · Cited by 22,179
Differential expression analysis for sequence count data
Anders, S., & Huber, W. (2010) · Genome Biology · doi:10.1186/gb-2010-11-10-r106 · Cited by 15,581
Trapnell, C., Williams, B. A., Pertea, G., Mortazavi, A., Kwan, G., van Baren, M. J., Salzberg, S. L., Wold, B. J., & Pachter, L. (2010) · Nature Biotechnology · doi:10.1038/nbt.1621 · Cited by 15,556
Haas, B. J., Papanicolaou, A., Yassour, M., Grabherr, M., Blood, P. D., Bowden, J., Couger, M. B., Eccles, D., Li, B., Lieber, M., et al. (2013) · Nature Protocols · doi:10.1038/nprot.2013.084 · Cited by 8,464
McCarthy, D. J., Chen, Y., & Smyth, G. K. (2012) · Nucleic Acids Research · doi:10.1093/nar/gks042 · Cited by 5,195
Metadata
- Authors
- Bullard, J. H., Purdom, E., Hansen, K. D., & Dudoit, S.
- Affiliations
- University of California, Berkeley (US)
- Corresponding author
- Bullard, James H
- Venue
- BMC Bioinformatics
- Published
- 2010-02-18
- Volume
- 11
- Issue
- 1
- Pages
- 94-94
- ISSN
- 1471-2105 (ISSN-L)
- PMC
- PMC2838869
- OpenAlex
- W2156631105
- Cited by
- 1,680
- References
- 23
- Citations by year
- 2026: 29 · 2025: 51 · 2024: 69 · 2023: 68 · 2022: 92 · 2021: 132 · 2020: 127 · 2019: 123 · 2018: 123 · 2017: 133 · 2016: 121 · 2015: 129 · 2014: 161 · 2013: 128 · 2012: 98
- FWCI
- 41.79
- Type
- journal article
- Language
- en
- Open access
- gold
- doi.org
- Licence
- Locations
- BMC Bioinformatics (publishedVersion, cc-by)
- PubMed Central (publishedVersion, cc-by)
- PubMed (publishedVersion)
- Collection of Biostatistics Research Archive (submittedVersion)
- DOAJ (DOAJ: Directory of Open Access Journals) (submittedVersion, cc-by-sa)
- eScholarship (California Digital Library) (submittedVersion)
- Funders
- National Science Foundation
- National Institutes of Health
- National Human Genome Research Institute (U01HG004271)
- Topics
- Life Sciences › Biochemistry, Genetics and Molecular Biology › Molecular Biology › Gene expression and cancer classification
- Life Sciences › Biochemistry, Genetics and Molecular Biology › Molecular Biology › Genomics and Phylogenetic Studies
- Life Sciences › Biochemistry, Genetics and Molecular Biology › Molecular Biology › Single-cell and spatial transcriptomics
- Keywords
- mRNA-seq, read count normalization, differential gene expression, method comparison
- MeSH terms
- RNA, Messenger, Sequence Analysis, RNA, Computational Biology, Databases, Genetic
About these checks
Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.
Statistics and arithmetic are recomputed, not judged.
Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.
What counts as a claim ›How we measure confidence ›How we check this ›
Are you an author, or do you see a problem with this page? Request a review
