Using Poisson mixed-effects model to quantify transcript-level gene expression in RNA-Seq

Abstract

Motivation: RNA sequencing (RNA-Seq) is a powerful new technology for mapping and quantifying transcriptomes using ultra high-throughput next generation sequencing technologies. Using deep sequencing, gene expression levels of all transcripts including novel ones can be quantified digitally. Although extremely promising, the massive amounts of data generated by RNA-Seq, substantial biases, and uncertainty in short read alignment pose challenges for data analysis. In particular, large base-specific variation and between-base dependence make simple approaches, such as those that use averaging to normalize RNA-Seq data and quantify gene expressions, ineffective. Results: In this study, we propose a Poisson mixed-effects (or in short, POME) model to characterize base-level read coverage within each transcript. The underlying expression level is included as a key parameter in this model. Because the proposed model is capable of incorporating base-specific variation as well as between-base dependence that affect read coverage profile throughout the transcript, it can lead to improved quantification of the true underlying expression level. Availability and Implementation: POME can be freely downloaded at http://www.stat.purdue.edu/~yuzhu/pome.html.