TL;DR
Researchers have launched RNAInter, a significantly expanded and enhanced repository for RNA-associated interactions, now encompassing over 41 million entries across 154 species. This update addresses the growing complexity of RNA biology by integrating diverse interaction types (RNA-RNA, RNA-protein, RNA-DNA, RNA-compound) and providing richer annotations, advanced search functionalities, and embedded analysis tools to facilitate a systems-level understanding of cellular RNA functions.
What Problem Does This Paper Address?
The study of RNA interactions has exploded, moving beyond simple RNA-RNA and RNA-protein pairings to include RNA-DNA and RNA-compound interactions. This burgeoning field generates vast amounts of data, but a significant challenge has been the lack of a centralized, comprehensive resource that integrates these diverse interaction types with detailed annotations. Existing databases often focus on specific interaction types or species, and many lack the advanced features needed for in-depth analysis. This fragmentation hinders researchers from gaining a holistic view of the "RNA interactome" and understanding the complex regulatory landscape of cellular RNAs. The authors aim to bridge this gap by updating and expanding their previous resource, RAID, into RNAInter, a more powerful and encompassing platform.
Methods
RNAInter was built by integrating data from a multitude of sources, including manual curation of literature and the incorporation of 35 existing databases. The data collection process involved screening PubMed for relevant keywords and then systematically processing the retrieved studies. For data organization, they unified various interactor symbols (miRNA, DNA, protein, compound, histone modification) and enriched the annotations with information such as RNA editing, localization, modification, structure, and homology interactions. They also incorporated tools for predicting RNA secondary structure. To visualize dynamic RNA expression, they utilized single-cell RNA-sequencing data, processing it to normalize expression levels and evaluate correlations.
Datasets
RNAInter integrates data from over 31,000 published studies, encompassing 419,522 RNA-associated interactions. Additionally, it incorporates data from 24 experimentally validated databases and 14 computationally predicted databases. For dynamic expression analysis, they utilized single-cell RNA-sequencing data from various sources, including human bone marrow cells (GSE75478), mouse precursor-hematopoietic stem cells (GSE67120), human testicular cells (GSE106487), and mouse spermatogenic cells (GSE107644).
Key Insights and Ideas
- The RNA Interactome is Vast and Complex: The sheer scale of RNA-associated interactions (over 41 million entries) highlights the central role of RNA in cellular regulation, extending far beyond its traditional roles.
- Integration is Key: By bringing together diverse interaction types (RNA-RNA, RNA-protein, RNA-DNA, RNA-compound) and data from numerous sources, RNAInter provides a much-needed unified view.
- Richer Annotations Drive Deeper Understanding: Going beyond simple interaction partners, detailed annotations on RNA editing, localization, modification, and structure are crucial for deciphering functional mechanisms.
- Tools Enhance Usability: Embedding analysis tools like RIscoper, IntaRNA, PRIdictor, and DeepBind directly within the database empowers users to perform immediate, sophisticated analyses without leaving the platform.
- Dynamic Expression is Critical: Visualizing RNA expression patterns across different cell stages and conditions offers dynamic insights into how RNA interactions change and contribute to biological processes.
Key Results
RNAInter now contains over 41 million RNA-associated interaction entries, a significant increase from its predecessor RAID v2.0. This includes:
- Over 34 million RNA-protein interactions (RPIs).
- Over 6 million RNA-RNA interactions (RRIs).
- Over 1 million RNA-histone modification interactions (RHIs).
- Over 138,000 RNA-DNA interactions (RDIs).
- Over 8,800 RNA-compound interactions (RCIs).
The database covers 154 species, an increase from 60 in RAID v2.0, and includes 22 RNA types, with eight new categories added. The platform offers advanced search functionalities (fuzzy and batch search), interaction network visualization, RNA dynamic expression plotting, and four integrated analysis tools.
Limitations and Future Work
The authors acknowledge that while RNAInter is comprehensive, it is a continuously evolving resource. Future work will focus on:
- Continuous Data Updates: Regularly updating manual curation of literature and integrating new high-throughput data.
- Expanding Annotations: Continuously adding more information about RNAs and other interacting molecules.
- Improving Confidence Scores: Optimizing the confidence score strategy as new sequencing technologies and prediction algorithms emerge.
- Integrating More Applications: Docking additional RNA-associated applications with the database.
- Enhancing Dynamic Expression Visualization: Providing more precise depictions of dynamic expression patterns, particularly from single-cell RNA-sequencing data.
- Optimizing Search and Visualization: Further refining search capabilities and visualization tools to present the most comprehensive map of the RNA interactome.
Summary updated