Sample size calculation while controlling false discovery rate for differential expression analysis with RNA-sequencing experiments

Abstract

BackgroundRNA-Sequencing (RNA-seq) experiments have been popularly applied to transcriptome studies in recent years. Such experiments are still relatively costly. As a result, RNA-seq experiments often employ a small number of replicates. Power analysis and sample size calculation are challenging in the context of differential expression analysis with RNA-seq data. One challenge is that there are no closed-form formulae to calculate power for the popularly applied tests for differential expression analysis. In addition, false discovery rate (FDR), instead of family-wise type I error rate, is controlled for the multiple testing error in RNA-seq data analysis. So far, there are very few proposals on sample size calculation for RNA-seq experiments.ResultsIn this paper, we propose a procedure for sample size calculation while controlling FDR for RNA-seq experimental design. Our procedure is based on the weighted linear model analysis facilitated by the voom method which has been shown to have competitive performance in terms of power and FDR control for RNA-seq differential expression analysis. We derive a method that approximates the average power across the differentially expressed genes, and then calculate the sample size to achieve a desired average power while controlling FDR. Simulation results demonstrate that the actual power of several popularly applied tests for differential expression is achieved and is close to the desired power for RNA-seq data with sample size calculated based on our method.ConclusionsOur proposed method provides an efficient algorithm to calculate sample size while controlling FDR for RNA-seq experimental design. We also provide an R package ssizeRNA that implements our proposed method and can be downloaded from the Comprehensive R Archive Network (http://cran.r-project.org).Electronic supplementary materialThe online version of this article (doi:10.1186/s12859-016-0994-9) contains supplementary material, which is available to authorized users.

Transcriptome sequencing identified hub genes for hepatocellular carcinoma by weighted-gene co-expression analysis

“…We estimated the statistical power for detecting significantly DEGs using ‘RnaSeqSampleSize’ [25] and ‘ssizeRNA’ [26]. The genes with minimun read counts > 10 across all individuals ( n = 11,076) genes was used.…”

“…Given a minimal FC of 2 (i.e., the effect size) and a FDR < 0.05, the statistical power to reject the null hypothesis that the population means of the two groups are equal is 0.697 (by an exact test [27]) using ‘RnaSeqSampleSize’ [25]. Given the same effect size and significance level, the achieved statistical power is 0.265 in nine pairs HCCs (by a paired t -test) using ‘ssizeRNA’ [26]. …”

Gene regulatory divergence between locally adapted ecotypes in their native habitats

“…Because the inland site had one extra plant per genotype compared with the coastal site, we analysed data from 12 randomly chosen parental lines of each genotype to equalize the power to detect effects between sites. A post hoc power analysis was conducted using the R package RnaSeqSampleSize (Bi & Liu, ).…”

Transcriptome Analysis of Cadmium-Treated Roots in Maize (Zea mays L.)

“…The unigene expression levels were calculated using reads per kilobase per million reads (RPKM), which eliminated the influences of gene length and sequencing level during the calculation of gene expression. A general Chi-squared test of statistical significance was used, and the false discovery rate (FDR) for the results was controlled (FDR < 0.05) (Ran and Peng, 2016). Significantly altered genes were described using heatmap analysis with unsupervised hierarchical clustering.…”