MntJULiP and Jutils: differential splicing analysis of RNA-seq data with covariates

TL;DR

This paper introduces new tools for analyzing RNA-seq data that account for factors like age and sex, improving accuracy in detecting splicing differences.

Contribution

The novel contribution is extending MntJULiP and Jutils to handle covariates, enabling more precise differential splicing analysis.

Findings

MntJULiP achieves high precision (>90%) by adjusting for covariates, reducing false positives.

Analysis of GTEx brain data reveals splicing differences increase with age group distance.

Covariate-adjusted clustering identifies a subgroup with distinct splicing patterns over time.

Abstract

Emerging large and complex RNA-seq datasets from disease and population studies include multiple confounders such as sex, age, ethnicity, and clinical attributes, which demand highly specialized data analysis tools. However, current methods are generally not equipped to handle the new challenges. We describe an extension of our programs MntJULiP and Jutils for differential splicing detection and visualization from RNA-seq data that accounts for covariates. MntJULiP detects intron-level differences in both splicing ratios and splicing abundance from RNA-seq data using a Bayesian linear mixture model adjusted for covariates. Jutils visualizes alternative variation with heatmaps, sashimi plots, Venn diagrams, and, reported here, PCA maps. With covariate modeling, MntJULiP drastically reduces false positives to achieve very high precision (>90%), significantly outperforming competitors. We applied the methods to GTEx brain RNA-seq samples to deconvolute the effects of sex and age at death on the splicing patterns. In particular, analyses of frontal cortex data reveal a pattern of increased splicing differences with more distant age groups, while clustering of covariate-adjusted data identifies a subgroup of individuals undergoing a distinct splicing program over the age span.

Figures