CoChat
Sign Up
Cited by 224Single-cell and spatial transcriptomicsGene expression and cancer classificationGenomics and Phylogenetic Studies

Separating measurement and expression models clarifies confusion in single-cell RNA sequencing analysis

Open access
Sarkar, Abhishek, Stephens, Matthew·2021·Nature Genetics
DOI 10.1038/s41588-021-00873-4PMC PMC8370014

Published version of a preprint

Open this paper in CoChat — free accountOpen PDF

Abstract

The high proportion of zeros in typical single-cell RNA sequencing datasets has led to widespread but inconsistent use of terminology such as dropout and missing data. Here, we argue that much of this terminology is unhelpful and confusing, and outline simple ideas to help to reduce confusion. These include: (1) observed single-cell RNA sequencing counts reflect both true gene expression levels and measurement error, and carefully distinguishing between these contributions helps to clarify thinking; and (2) method development should start with a Poisson measurement model, rather than more complex models, because it is simple and generally consistent with existing data. We outline how several existing methods can be viewed within this framework and highlight how these methods differ in their assumptions about expression variation. We also illustrate how our perspective helps to address questions of biological interest, such as whether messenger RNA expression levels are multimodal among cells.

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.

A few free checks a day · papers already verified open instantly · results are public

Integrity signals

Registry

Crossref · 10.1038/s41588-021-00873-4

Cites

12 of 12 references verified0 citation errors · 0 not checkable

Statistics

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.

A few free checks a day · papers already verified open instantly · results are public

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?

  • Sequencing coverage results in some dropout (apparent zeroes) for each gene, and scRNA-seq is conventionally used to characterize cell types and pathways governed by changes in multiple transcripts.

    “Because sequencing coverage results in some dropout (apparent zeroes) for each gene, scRNA-seq is conventionally used to characterize cell types and pathways, processes governed by changes in multiple transcripts [50].”

    From Elucidating brain transport pathways and cell type-dependent gene silencing of a durable lipid–siRNA conjugate administered into cerebrospinal fluid (2025) · cites this paper as [50]

    The Introduction describes scRNA-seq as a way to investigate cell types, cellular processes, and pathways. The Modeling scRNA-seq measurement section says lower sequencing depth contributes to the high proportion of zeros, but also identifies low or truly zero expression as a cause; it does not say that coverage causes dropout for each gene. The paper also cautions in A call to simplify terminology that “dropout” is ambiguous and recommends calling these observations zeros. (LLM, conf=0.87)

    Experimental — accuracy not yet measured

Self-check

Percentages, sums and reported differences re-computed from the paper’s own numbers.

The self-check needs the paper’s full text.

Cites

Flagged first, then by citations

Research works printed but not linked: 48

  1. [1]Fuller WA Measurement Error Models (John Wiley & Sons, Inc., USA, 1986).
  2. [2]Marioni JC, Mason CE, Mane SM, Stephens M & Gilad Y RNA-seq: An assessment of technical reproducibility and comparison with gene expression arrays. Genome Res 18, 1509– 1517 (2008). 18550803 PMC2527709
  3. [3]Pachter L Models for transcript quantification from RNA-Seq. arXiv e-prints (2011).
  4. [4]Wang J Gene expression distribution deconvolution in single-cell RNA sequencing. Proc Natl Acad Sci USA (2018). PMC6048536 29946020
  5. [6]Kharchenko PV, Silberstein L & Scadden DT Bayesian approach to single-cell differential expression analysis. Nat Methods 11, 740 (2014). 24836921 PMC4112276
  6. [8]Haque A, Engel J, Teichmann SA & Lönnberg T A practical guide to single-cell RNA-sequencing for biomedical research and clinical applications. Genome Med 9, 75 (2017). 28821273 PMC5561556
  7. [9]Zhu L, Lei J, Devlin B & Roeder K A unified statistical framework for single cell and bulk RNA sequencing data. Ann Appl Stat 12, 609– 632 (2018). 30174778 PMC6114100
  8. [10]Qiu P Embracing the dropouts in single-cell RNA-seq analysis. Nat Commun 11, 1169 (2020). 32127540 PMC7054558
  9. [11]Fujimura FK, Northrup H, Beaudet AL & O’Brien WE Genotyping Errors with the Polymerase Chain Reaction. N Engl J Med 322, 61– 61 (1990). 2294422
  10. [12]Whale AS, Cowen S, Foy CA & Huggett JF Methods for Applying Accurate Digital PCR Analysis on Low Copy DNA Samples. PLoS ONE 8, 1– 10 (2013). PMC3589384 23472156
  11. [13]Hicks SC, Townes FW, Teng M & Irizarry RA Missing data and technical variability in single-cell RNA-sequencing experiments. Biostatistics (2017). PMC6215955 29121214
  12. [14]Li WV & Li JJ An accurate and robust imputation method scImpute for single-cell RNA-seq data. Nat Commun 9, 997 (2018). 29520097 PMC5843666
  13. [15]Chen M & Zhou X VIPER: variability-preserving imputation for accurate gene expression recovery in single-cell RNA sequencing studies. Genome Biol 19, 196 (2018). 30419955 PMC6233584
  14. [16]Talwar D, Mongia A, Sengupta D & Majumdar A AutoImpute: Autoencoder based imputation of single-cell RNA-seq data. Sci Rep 8, 16329 (2018). 30397240 PMC6218547
  15. [17]Svensson V Droplet scRNA-seq is not zero-inflated. Nat Biotech (2020). 31937974
  16. [20]Huang M SAVER: gene expression recovery for single-cell RNA sequencing. Nat Methods 15, 539– 542 (2018). 29941873 PMC6030502
  17. [21]Eraslan G, Simon LM, Mircea M, Mueller NS & Theis FJ Single-cell RNA-seq denoising using a deep count autoencoder. Nat Commun 10, 390 (2019). 30674886 PMC6344535
  18. [22]Tang W bayNorm: Bayesian gene expression recovery, imputation and normalization for single-cell RNA-sequencing data. Bioinformatics 36, 1174– 1181 (2019). PMC7703772 31584606
  19. [23]Hilbe JM Modeling Count Data (Cambridge University Press, 2014).
  20. [24]Lu M Generalized Adaptive Shrinkage Methods and Applications in Genomics Studies PhD thesis (University of Chicago, 2018).
  21. [25]Raj A & van Oudenaarden A Nature, Nurture, or Chance: Stochastic Gene Expression and Its Consequences. Cell 135, 216– 226 (2008). 18957198 PMC3118044
  22. [26]Shalek AK Single-cell transcriptomics reveals bimodality in expression and splicing in immune cells. Nature 498, 236 (2013). 23685454 PMC3683364
  23. [27]Shalek AK Single-cell RNA-seq reveals dynamic paracrine control of cellular variation. Nature 510, 363 (2014). 24919153 PMC4193940
  24. [28]Bacher R & Kendziorski C Design and computational analysis of single-cell RNA-sequencing experiments. Genome Biol 17, 63 (2016). 27052890 PMC4823857
  25. [30]Hu Q & Greene CS Parameter tuning is a key part of dimensionality reduction via deep variational autoencoders for single cell RNA transcriptomics. Pac Symp Bio-comput 24, 362– 373 (2019). PMC6417816 30963075
  26. [32]Brennecke P Accounting for technical noise in single-cell RNA-seq experiments. Nat Methods 10, 1093 (2013). 24056876
  27. [34]Wang W & Stephens M Empirical Bayes Matrix Factorization. Journal of Machine Learning Research 22 (in press). PMC10621241 37920532
  28. [35]Tipping ME & Bishop CM Probabilistic Principal Component Analysis. Journal of the Royal Statistical Society: Series B (Statistical Methodology) 61, 611– 622.
  29. [36]Pierson E & Yau C ZIFA: Dimensionality reduction for zero-inflated single-cell gene expression analysis. Genome Biol 16, 241 (2015). 26527291 PMC4630968
  30. [38]Verma A & Engelhardt BE A robust nonlinear low-dimensional manifold for single cell RNA-seq data. BMC Bioinformatics 21, 324 (2020). 32693778 PMC7374962
  31. [39]Law CW, Chen Y, Shi W & Smyth GK voom: precision weights unlock linear model analysis tools for RNA-seq read counts. Genome Biol 15, R29 (2014). 24485249 PMC4053721
  32. [40]Lun A Overcoming systematic errors caused by log-transformation of normalized single-cell RNA sequencing data. bioRxiv (2018).
  33. [41]Venables WN & Ripley BD Modern Applied Statistics with S Fourth. ISBN 0-387-95457-0 (Springer, New York, 2002).
  34. [42]Robinson MD, McCarthy DJ & Smyth GK edgeR: a Bioconductor package for differential expression analysis of digital gene expression data. Bioinformatics 26, 139– 140 (2009). 19910308 PMC2796818
  35. [43]Love MI, Huber W & Anders S Moderated estimation of fold change and dispersion for RNA-seq data with DESeq2. Genome Biol 15, 550 (2014). 25516281 PMC4302049
  36. [45]Zeileis A, Kleiber C & Jackman S Regression Models for Count Data in R. Journal of Statistical Software 27 (2008).
  37. [46]Stephens M False discovery rates: a new deal. Biostatistics 18, 275– 294 (2017). 27756721 PMC5379932
  38. [48]Lee DD & Seung HS Algorithms for Non-negative Matrix Factorization in Advances in Neural Information Processing Systems 13, Papers from Neural Information Processing Systems (NIPS) 2000, Denver, CO, USA (eds Leen TK, Dietterich TG & Tresp V) (MIT Press, 2000), 556– 562.
  39. [49]Levitin HM De novo gene signature identification from single-cell RNA-seq with hierarchical Poisson factorization. Mol Syst Biol 15 (2019). PMC6386217 30796088
  40. [50]Gouvert O, Oberlin T & Févotte C Negative Binomial Matrix Factorization. IEEE Signal Processing Letters 27, 815– 819 (2020).
  41. [51]Sun S, Chen Y, Liu Y & Shang X A fast and efficient count-based matrix factorization method for detecting cell types from single-cell RNAseq data. BMC Syst Biol 13, 28 (2019). 30953530 PMC6449882
  42. [52]Risso D, Perraudeau F, Gribkova S, Dudoit S & Vert J-P A general and flexible method for signal extraction from single-cell RNA-seq data. Nat Commun 9, 284 (2018). 29348443 PMC5773593
  43. [54]Sarkar AK Discovery and characterization of variance QTLs in human induced pluripotent stem cells. PLoS Genetics 15, 1– 16 (2019). PMC6474585 31002671
  44. [56]Stewart BJ Spatiotemporal immune zonation of the human kidney. Science 365, 1461– 1466 (2019). 31604275 PMC7343525
  45. [57]Lukowski SW A single-cell transcriptome atlas of the adult human retina. The EMBO Journal 38, e100811 (2019). 31436334 PMC6745503
  46. [59]Macosko EZ Highly Parallel Genome-wide Expression Profiling of Individual Cells Using Nanoliter Droplets. Cell 161, 1202– 1214 (2015). 26000488 PMC4481139
  47. [60]Klein AM Droplet Barcoding for Single-Cell Transcriptomics Applied to Embryonic Stem Cells. Cell 161, 1187– 1201 (2015). 26000487 PMC4441768
  48. [61]Islam S Characterization of the single-cell transcriptional landscape by highly multiplex RNA-seq. Genome Res 21, 1160– 1167 (2011). 21543516 PMC3129258

Cited by

Sorted by citations

Authors
Sarkar, A., & Stephens, M.
Affiliations
University of Chicago (US)
Corresponding author
  • Sarkar, Abhishek
  • Stephens, Matthew
ORCID
Venue
Nature Genetics
Publisher
Springer Science and Business Media LLC
Published
2021-05-24
Volume
53
Issue
6
Pages
770-777
ISSN
  • 1061-4036 (ISSN-L)
  • 1546-1718
DOI
10.1038/s41588-021-00873-4
PMC
PMC8370014
OpenAlex
W3164156820
Cited by
224
References
12
Citations by year
2026: 30 · 2025: 51 · 2024: 52 · 2023: 31 · 2022: 39 · 2021: 17 · 2020: 4
FWCI
4.02
Type
journal article
Language
en
Open access
Licence
Locations
  • Nature Genetics (publishedVersion)
  • bioRxiv (Cold Spring Harbor Laboratory) (acceptedVersion, cc-by)
  • PubMed (publishedVersion)
  • PubMed Central (acceptedVersion)
Funders
  • U.S. Department of Health & Human Services | National Institutes of Health (HG002585)
  • Leona M. and Harry B. Helmsley Charitable Trust (N/A)
Topics
  • Life Sciences › Biochemistry, Genetics and Molecular Biology › Molecular Biology › Single-cell and spatial transcriptomics
  • 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
Keywords
single-cell RNA sequencing, dropout, missing data, measurement error, gene expression, expression variation, count data, statistical modeling
MeSH terms
Gene Expression Regulation, Humans, Models, Genetic, Terminology as Topic, Poisson Distribution, Sequence Analysis, RNA, Single-Cell Analysis

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