DStruct: identifying differentially reactive regions from RNA structurome profiling data

Krishna Choudhary,Yu-Hsuan Lai,Sharon Aviran,Elizabeth J Tran

doi:10.1186/s13059-019-1641-3

Krishna Choudhary, Yu-Hsuan Lai + Show 2 more

Open Access

https://doi.org/10.1186/s13059-019-1641-3

Copy DOI

Abstract

RNA biology is revolutionized by recent developments of diverse high-throughput technologies for transcriptome-wide profiling of molecular RNA structures. RNA structurome profiling data can be used to identify differentially structured regions between groups of samples. Existing methods are limited in scope to specific technologies and/or do not account for biological variation. Here, we present dStruct which is the first broadly applicable method for differential analysis accounting for biological variation in structurome profiling data. dStruct is compatible with diverse profiling technologies, is validated with experimental data and simulations, and outperforms existing methods.

Highlights

RNA molecules adopt diverse and intricate structures, which confer on them the capacity to perform key roles in myriads of cellular processes [1, 2]
We described dStruct, a novel approach to identify Differentially reactive region (DRR) from structure profiling (SP) data. dStruct is compatible with diverse SP protocols and accounts for biological variation in SP data
It constructs regions that are potential candidates for DRRs to facilitate de novo discovery or utilizes candidate regions identified by collateral studies to aid guided discovery

Summary

Introduction

RNA molecules adopt diverse and intricate structures, which confer on them the capacity to perform key roles in myriads of cellular processes [1, 2]. Recent developments have led to a diversity of structure probing or structure profiling (SP) technologies [4, 5] These technologies have made it possible to perform comparative analysis of structures of select RNAs or whole RNA structuromes simultaneously [6,7,8,9,10,11,12,13,14,15,16,17,18,19,20,21]. SP technologies result in nucleotide-level scores, called as reactivities, that summarize one or more aspects of local structure (e.g., steric constraint due to base pairing interaction). To this end, they utilize probing reagents that react with RNA nucleotides in a structure-sensitive manner. A number of reagents (e.g., SHAPE, DMS, nucleases) exist, which react with sensitivity to different aspects of local stereochemistry [22,23,24]

Objectives

Methods

Results

Discussion

Conclusion