PAREameters: a tool for computational inference of plant miRNA-mRNA targeting rules using small RNA and degradome sequencing data.

Vincent Moulton,Joshua Thody,Irina Mohorianu

doi:10.1093/nar/gkz1234

Vincent Moulton, Joshua Thody + Show 1 more

Open Access

https://doi.org/10.1093/nar/gkz1234

Copy DOI

Abstract

MicroRNAs (miRNAs) are short, non-coding RNAs that modulate the translation-rate of messenger RNAs (mRNAs) by directing the RNA-induced silencing complex to sequence-specific targets. In plants, this typically results in cleavage and subsequent degradation of the mRNA. Degradome sequencing is a high-throughput technique developed to capture cleaved mRNA fragments and thus can be used to support miRNA target prediction. The current criteria used for miRNA target prediction were inferred on a limited number of experimentally validated A. thaliana interactions and were adapted to fit these specific interactions; thus, these fixed criteria may not be optimal across all datasets (organisms, tissues or treatments). We present a new tool, PAREameters, for inferring targeting criteria from small RNA and degradome sequencing datasets. We evaluate its performance using a more extensive set of experimentally validated interactions in multiple A. thaliana datasets. We also perform comprehensive analyses to highlight and quantify the differences between subsets of miRNA–mRNA interactions in model and non-model organisms. Our results show increased sensitivity in A. thaliana when using the PAREameters inferred criteria and that using data-driven criteria enables the identification of additional interactions that further our understanding of the RNA silencing pathway in both model and non-model organisms.

Highlights

Improvements to Generation Sequencing technologies have resulted in larger and more diverse experiments, including ones that make use of multiple data types, for example, to increase prediction accuracy of regulatory interactions by combining small RNA sequencing and messenger RNA quantification [1]
Small RNAs are short, non-coding RNAs with important roles in transcriptional and post-transcriptional gene regulation in eukaryotes [3]. The latter mode of action is achieved via a class of small RNA (sRNA), the microRNAs, which reduce the amount of messenger RNAs (mRNAs) available for translation by directing the RNA-induced silencing complex (RISC) to their sequence-specific mRNA target(s) and inducing cleavage and subsequent degradation of the mRNA [4]
We show that different subsets of miRNA– mRNA interactions, such as those containing conserved or species-specific miRNAs, those found in monocots and dicots, and those identified in model and non-model organism, display variation in their target interaction properties

Summary

Introduction

Improvements to Generation Sequencing technologies have resulted in larger and more diverse experiments, including ones that make use of multiple data types, for example, to increase prediction accuracy of regulatory interactions by combining small RNA (sRNA) sequencing and messenger RNA (mRNA) quantification [1]. These improvements have led to the sequencing and annotation of different organisms’ genomes and facilitated functional studies outside of the context of model organisms [2]. We investigate the applicability and portability of the current miRNA target interaction model

Methods

Results

Conclusion