Skip to main page content
U.S. flag

An official website of the United States government

Dot gov

The .gov means it’s official.
Federal government websites often end in .gov or .mil. Before sharing sensitive information, make sure you’re on a federal government site.

Https

The site is secure.
The https:// ensures that you are connecting to the official website and that any information you provide is encrypted and transmitted securely.

Access keys NCBI Homepage MyNCBI Homepage Main Content Main Navigation
. 2014 Feb 1;30(3):301-4.
doi: 10.1093/bioinformatics/btt688. Epub 2013 Dec 6.

RNA-seq differential expression studies: more sequence or more replication?

Yuwen Liu  1 , Jie Zhou, Kevin P White
Affiliations

RNA-seq differential expression studies: more sequence or more replication?

Yuwen Liu et al. Bioinformatics. .

Abstract

Motivation: RNA-seq is replacing microarrays as the primary tool for gene expression studies. Many RNA-seq studies have used insufficient biological replicates, resulting in low statistical power and inefficient use of sequencing resources.

Results: We show the explicit trade-off between more biological replicates and deeper sequencing in increasing power to detect differentially expressed (DE) genes. In the human cell line MCF7, adding more sequencing depth after 10 M reads gives diminishing returns on power to detect DE genes, whereas adding biological replicates improves power significantly regardless of sequencing depth. We also propose a cost-effectiveness metric for guiding the design of large-scale RNA-seq DE studies. Our analysis showed that sequencing less reads and performing more biological replication is an effective strategy to increase power and accuracy in large-scale differential expression RNA-seq studies, and provided new insights into efficient experiment design of RNA-seq studies.

Availability and implementation: The code used in this paper is provided on: http://home.uchicago.edu/∼jiezhou/replication/. The expression data is deposited in the Gene Expression Omnibus under the accession ID GSE51403.

PubMed Disclaimer

Figures

Fig. 1.
Fig. 1.
(a) Increase in biological replication significantly increases the number of DE genes identified. Numbers of sequencing reads have a diminishing return after 10 M reads. Line thickness indicates depth of replication, with 2 replicates the darkest and 7 replicates the lightest. The lines are smoothed averages for each replication level, with the shaded regions corresponding to the 95% confidence intervals. (b) Power of detecting DE genes increases with both sequencing depth and biological replication. Similar to the trends in (a), increases in the power showed diminishing returns after 10 M reads. (c) ROC curves for three biological replicates. Sequencing deeper than 10 M reads did not significantly improve statistical power and precision for detecting DE genes. (d) The CV of logFC for the top 100 DE genes. The CV of the logFC estimates decreased significantly as we added more biological replicates, whereas adding sequencing depth after 10 M reads had much less effect
Fig. 2.
Fig. 2.
(a–c) The CV of logCPM for high expression level genes (a), medium expression level genes (b) and low expression level genes (c) (see Section 2 for definition). High/medium expression level genes have low CV for expression level estimates. Adding sequencing depth did not have significant effect on accuracy of estimation, whereas adding biological replicates improved accuracy significantly. For low expression level genes, both adding sequencing depth and adding biological replication level improved expression level estimation accuracy. (d) Number of DE genes plotted against the total estimated sequencing cost. If higher numbers of DE genes are needed, increased biological replication should be used
See this image and copyright information in PMC

Similar articles

See all similar articles

Cited by

See all "Cited by" articles

References

    1. Anders S, Huber W. Differential expression analysis for sequence count data. Genome Biol. 2010;11:R106. - PMC - PubMed
    1. Auer PL, Doerge RW. Statistical design and analysis of RNA sequencing data. Genetics. 2010;185:405–416. - PMC - PubMed
    1. Brawand D, et al. The evolution of gene expression levels in mammalian organs. Nature. 2011;478:343–348. - PubMed
    1. Busby MA, et al. Scotty: a web tool for designing RNA-Seq experiments to measure differential gene expression. Bioinformatics. 2013;29:656–657. - PMC - PubMed
    1. Fang Z, Cui X. Design and validation issues in RNA-seq experiments. Brief. Bioinform. 2011;12:280–287. - PubMed

Publication types