Unravelling biomolecular and community grammars of RNA granules via machine learning

Abstract

AbstractRNA granules are dynamic compartments within cells that play a crucial role in posttranscriptional regulation of gene expression. They are associated with a variety of human neurodegenerative diseases. While RNA granules play vital roles in cellular functions, the comprehension of their assembly has remained elusive.In this study, we employed robust machine learning models combining residue content and physicochemical features to accurately identify potential RNA granule (i.e.,stress granule and P-body) proteome within the human proteome. Our models achieved good performance with high areas under the receiver operating characteristic curve of up to 0.88, outperforming previous liquid-liquid phase separation models. Intriguingly, the predicted RNA granule proteome reveals a significant enrichment in biological functions and domains associated with RNA granule-related processes, mirroring findings from observed high-confidence RNA granule protein datasets. Furthermore, our analysis unveils critical physicochemical attributes, notably hydrophobicity, influencing the formation of RNA granules.Using the constructed model, we uncovered the central roles of RNA granule proteins with high propensities within the comprehensive RNA granule protein-protein interaction (PPI) network and their commonality in diverse RNA granules. Furthermore, we identified prominent clusters with dense PPIs, significantly contributing to critical biological processes within diverse RNA granules, including translation, mRNA decay, rRNA processing, and mRNA splicing. This analysis proposes a hypothesis: dense PPI clusters are integral functional subunits, constituting relatively stable ‘cores’ within diverse RNA granules.In conclusion, this study provides a comprehensive molecular and community-based foundation for understanding the importance of PPIs in the stability of RNA granule formation and functionality. This analysis contributes to a deeper and more comprehensive understanding of the intricate nature of RNA granules and opens avenues for future research and therapeutic interventions targeting RNA granule- related diseases.