A mixed model approach for joint genetic analysis of alternatively spliced transcript isoforms using RNA-Seq data

TL;DR

This paper introduces a mixed model method for joint analysis of transcript isoforms from RNA-Seq data, enabling detailed genetic regulation mapping and accounting for variation sources, improving accuracy in understanding gene expression regulation.

Contribution

The paper presents a novel mixed model approach that jointly analyzes multiple transcript isoforms, accounting for various sources of variation, and maps isoform-specific genetic effects.

Findings

Improved calibration of statistical tests for genetic regulation.

Enhanced understanding of isoform-specific genetic effects.

Demonstrated effectiveness on human HapMap data.

Abstract

RNA-Seq technology allows for studying the transcriptional state of the cell at an unprecedented level of detail. Beyond quantification of whole-gene expression, it is now possible to disentangle the abundance of individual alternatively spliced transcript isoforms of a gene. A central question is to understand the regulatory processes that lead to differences in relative abundance variation due to external and genetic factors. Here, we present a mixed model approach that allows for (i) joint analysis and genetic mapping of multiple transcript isoforms and (ii) mapping of isoform-specific effects. Central to our approach is to comprehensively model the causes of variation and correlation between transcript isoforms, including the genomic background and technical quantification uncertainty. As a result, our method allows to accurately test for shared as well as transcript-specific genetic regulation of transcript isoforms and achieves substantially improved calibration of these statistical tests. Experiments on genotype and RNA-Seq data from 126 human HapMap individuals demonstrate that our model can help to obtain a more fine-grained picture of the genetic basis of gene expression variation.