Probe region expression estimation for RNA-seq data for improved microarray comparability

Abstract

Rapidly growing public gene expression databases contain a wealth of data for building an unprecedentedly detailed picture of human biology and disease. This data comes from many diverse measurement platforms that make integrating it all difficult. In this paper, we propose a new method for processing RNA-sequencing data that yields gene expression estimates that are much more similar to corresponding estimates from microarray data, hence greatly improving cross-platform comparability. The method we call PREBS is based on estimating the expression only from microarray probe regions, and processing these estimates with microarray summarisation algorithm RMA. This allows new ways of using RNA-sequencing data, such as expression estimation for microarray probe sets. Gene signatures defined based on PREBS expression measures of RNA-sequencing data are much more accurate for retrieval of similar microarray samples from a database.