AN INTEGRATIVE RNA-SEQ PIPELINE FOR LINKING MICRORNA AND MRNA DIFFERENTIAL EXPRESSION

Abstract

High-throughput sequencing technologies enable parallel measurement of small RNA and messenger RNA (mRNA) expression; however, most RNA-seq analysis workflows evaluate these data types independently. While differential expression analyses identify transcripts that change between conditions, they often do not directly address the regulatory relationships between microRNAs (miRNAs) and their predicted mRNA targets. This separation can limit biological interpretation and make it harder to identify meaningful relationships between miRNAs and their target genes. To address this challenge, RNA-integrate was developed as a reproducible and flexible RNA-seq workflow that combines DESeq2-based differential expression analysis of small RNA and mRNA expression data. The workflow is fully parameterized through external YAML configuration files and supports multiple RNA-seq quantification formats, allowing application across diverse datasets and experimental designs. Following differential expression analysis, miRNAs are linked to predicted target mRNAs using a user-defined gene-target table, generating paired datasets for integrative analysis. RNA-integrate produces a diverse set of outputs, including differential expression tables, quality-control visualizations, integrative results tables, interactive HTML reports, and publication-quality figures. Integrative analyses include cosmic plots and slope plots, which enable visualization of expression relationships between miRNAs and their predicted targets. In addition, the workflow implements a one-sided binomial sign test to evaluate whether inverse miRNA–mRNA relationships occur more frequently than expected by chance. To demonstrate the utility of the workflow, RNA-integrate was applied to a pash-1 mutant versus wild-type dataset in Caenorhabditis elegans. Because pash-1 is required for miRNA biogenesis, disruption of pash-1 provides a biologically relevant system for evaluating miRNA-mediated regulation. Analysis revealed a strong enrichment of inverse relationships among significant miRNA–mRNA target pairs, consistent with established models of post-transcriptional repression by miRNAs. By combining reproducible workflow design, differential expression analysis, integrative visualization, and statistical evaluation within a single framework, RNA-integrate provides an accessible platform for investigating RNA regulatory networks. The workflow facilitates cross-dataset analyses, improves reproducibility, and enables a flexible framework of miRNA–mRNA interactions across organisms and experimental systems.