A Multi-Layer, Multi-Species Graph-Based Framework for Comparative Analysis of sRNA-Mediated Regulation

Small RNAs (sRNAs) are central post-transcriptional regulators of bacterial gene expression, yet their evolutionary dynamics remain difficult to resolve due to limited sequence conservation. While sequence-homologous sRNAs can retain or diverge in function, unrelated sRNAs may independently evolve to regulate similar biological processes, making functional conservation invisible to sequence-based approaches. Here, we present a multi-layer, multi-species graph-based framework that integrates experimentally derived sRNA-mRNA interactions, orthology relationships, functional annotations, and semantic clustering of Gene Ontology biological process terms. The framework provides two complementary tools for assessing functional conservation among orthologous sRNAs and convergent regulation by distinct sRNAs across species, capturing functional signals that are weak in individual data sources, yet emerge from integrating multiple layers and species. Applied to six Gammaproteobacterial strains, our approach reveals that functional conservation is frequently maintained among orthologous sRNAs despite divergence in their target repertoires. We further identify lineage-specific regulatory modules and cases of convergent regulation, where non-homologous sRNAs from phylogenetically distant species independently regulate the same essential biological processes. These results demonstrate that sRNA-mediated regulation is both evolutionarily flexible and functionally constrained. The framework is modular, scalable, and adaptable to additional datasets and regulatory layers, establishing a generalizable and hypothesis-generating approach for studying the evolution of post-transcriptional regulatory networks across species.