Empirical Bayes analysis of RNA sequencing experiments with auxiliary information
The Annals of Applied Statistics · 2019 · Volume 13 · Issue 4
Abstract
Finding differentially expressed genes is a common task in high-throughput transcriptome studies. While traditional statistical methods rank the genes by their test statistics alone, we analyze an RNA sequencing dataset using the auxiliary information of gene length and the test statistics from a related microarray study. Given the auxiliary information, we propose a novel nonparametric empirical Bayes procedure to estimate the posterior probability of differential expression for each gene. We demonstrate the advantage of our procedure in extensive simulation studies and a psoriasis RNA sequencing study. The companion R package calm is available at Bioconductor.
Keywords
Conditional density, multiple comparison, Nonparametric regression, simultaneous inference
Enable JavaScript for citation tools and additional paper interactions.