- RNA structure databases are specialized biological resources that collect, organize, annotate, and distribute information about the structural organization of ribonucleic acid (RNA) molecules. They support research into the relationship between RNA sequence, structure, and function, helping scientists investigate gene regulation, protein synthesis, RNA processing, molecular recognition, and disease mechanisms. RNA molecules perform diverse biological roles, including carrying genetic information, catalyzing biochemical reactions, regulating gene expression, and guiding molecular complexes. Their functions often depend on their ability to fold into specific structural arrangements, making RNA structure databases essential tools in molecular biology, structural bioinformatics, genomics, and computational biology.
- RNA structure is commonly described at two major levels: secondary structure and tertiary structure. Secondary structure represents local interactions between nucleotides, particularly complementary base pairing, which forms structural elements such as stems, hairpin loops, internal loops, bulges, and junctions. Tertiary structure describes the complete three-dimensional arrangement of the RNA molecule, including long-range interactions, base stacking, non-canonical base pairs, and contacts between distant regions of the sequence. RNA structure databases may also contain information about pseudoknots, conserved motifs, chemical modifications, protein-binding sites, and alternative conformations. These different levels of structural information are important because an RNA molecule’s biological activity may depend on both its local folding and its overall three-dimensional organization.
- The Protein Data Bank (PDB) is a primary resource for experimentally determined three-dimensional structures of biological macromolecules, including RNA, DNA, proteins, and their complexes. RNA structures deposited in the PDB may have been determined using X-ray crystallography, nuclear magnetic resonance spectroscopy, cryo-electron microscopy, or other structural methods. Records can contain atomic coordinates, molecular sequences, experimental metadata, structural assemblies, bound ligands, and validation information. Researchers use PDB structures to investigate ribosomes, transfer RNA, ribozymes, riboswitches, spliceosomal complexes, viral RNA, and RNA–protein interactions. Because many RNA molecules function as parts of larger assemblies, the biological context and interactions with proteins or other molecules are often essential when interpreting their structures.
- The Nucleic Acid Knowledgebase (NAKB), formerly known as the Nucleic Acid Database (NDB), is a specialized resource for nucleic acid structural information and analysis. It provides access to structural data and annotations for DNA- and RNA-containing molecules, together with tools for examining nucleic acid geometry, base pairing, and structural features. Its associated resources include structural annotation software such as 3DNA/DSSR and RNAView, which can help identify base pairs, base stacking, and other features in experimentally determined structures. NAKB also integrates information from structural resources and provides representative RNA structure sets and related annotations. These capabilities are useful for detailed analyses of RNA architecture and for comparing structural features across different molecules. <Cite refs={[“turn547691search2″,”turn547691search13”]}/>
- RNA 3D Hub, developed by the RNA structural bioinformatics community at Bowling Green State University, is another important resource for analyzing experimentally determined RNA structures. It organizes RNA-containing three-dimensional structures and provides structural annotations, representative sets, and collections of recurring structural motifs. Its RNA Structure Atlas includes annotations of pairwise nucleotide interactions and structural elements such as hairpin loops, internal loops, and three-way junctions. The RNA 3D Motif Atlas collects representative examples of recurring RNA motifs, while representative structure sets group related structures to reduce redundancy. These resources help researchers compare RNA folds, examine conserved structural elements, and investigate how sequence variation relates to structural organization. Tools such as WebFR3D support searches for RNA three-dimensional motifs, while JAR3D helps identify structural motifs in RNA sequences and alignments. <Cite refs={[“turn547691search7″,”turn547691search10”]}/>
- RNAcentral is an integrated resource for non-coding RNA sequences from numerous expert databases. Although it is primarily a sequence database rather than a dedicated archive of experimentally determined three-dimensional structures, it is highly relevant to RNA structural bioinformatics because it provides secondary-structure diagrams for many RNA entries and links sequences to structural and functional annotations. RNAcentral integrates data from resources covering microRNAs, transfer RNAs, ribosomal RNAs, long non-coding RNAs, small nucleolar RNAs, and other RNA types. It assigns stable identifiers to distinct RNA sequences and supports text searches, sequence similarity searches, bulk downloads, and programmatic access. Its integration with other databases allows researchers to connect RNA sequences with conserved families, genomic locations, annotations, and available structural information. <Cite refs={[“turn547691search6″,”turn547691search8”]}/>
- Rfam is a database of RNA families that uses sequence alignments, consensus secondary structures, and covariance models to represent conserved RNA families and structural patterns. It includes families of non-coding RNAs and structured RNA elements, such as riboswitches, ribosomal RNAs, transfer RNAs, and regulatory RNA motifs. Covariance models account for sequence conservation and compensatory changes in paired positions, making them useful for identifying RNA families whose members retain similar structures despite sequence differences. Rfam can therefore support RNA family classification, homology searches, genome annotation, and the identification of structured RNA elements in newly sequenced organisms. Its family alignments and consensus structures can also provide a framework for investigating the evolution and functional importance of RNA secondary structures. <Cite refs={[“turn547691search0″,”turn547691search4”]}/>
- The RNA STRAND database, historically known as a collection of RNA secondary structures, provides examples of experimentally determined and computationally annotated RNA secondary structures. Such datasets have been used to develop and evaluate RNA folding algorithms, compare predicted structures with reference structures, and benchmark computational methods. Secondary-structure datasets can represent structures in formats such as dot-bracket notation, connectivity tables, or base-pair lists. Researchers should check the availability, provenance, release status, and licensing of a particular RNA STRAND dataset before using it, because database maintenance and access arrangements may change over time. For current RNA secondary-structure resources, RNAcentral and Rfam can also provide useful structural annotations and family-level information.
- RNA structure databases increasingly include computationally predicted structures in addition to experimentally determined models. The NuFold RNA Structure Database, for example, provides predicted three-dimensional RNA models generated using the NuFold deep-learning approach. Its records include structural visualizations and confidence-related information intended to help researchers assess predictions. Such resources can support the investigation of RNA molecules for which experimental structures are unavailable and can generate hypotheses for further study. However, predicted structures must be distinguished clearly from experimentally determined structures, and their reliability should be assessed using the relevant confidence metrics, structural validation methods, and independent biological evidence. <Cite refs={[“turn547691search3”]}/>
- Other specialized resources focus on RNA structural diversity and recurring molecular interactions. CoDNaS-RNA collects alternative experimentally observed conformations of RNA molecules, allowing researchers to investigate structural variation associated with different experimental conditions, molecular environments, and interactions. The resource is useful for examining the fact that RNA does not always exist as one fixed structure: a molecule may adopt multiple conformations that contribute to its biological activity. The RNA 3D Motif Atlas and related RNA 3D Hub resources support the identification and comparison of conserved motifs, while RNA Junction Database and related tools provide information about structural junctions, kissing loops, and bulges. These resources are particularly valuable for studies of RNA folding, molecular recognition, and the structural basis of RNA function. <Cite refs={[“turn547691search11″,”turn547691search7″,”turn547691search1”]}/>
- RNA structure databases support a wide range of bioinformatics applications. Researchers can use them to study RNA folding, identify conserved structural motifs, annotate non-coding RNA genes, investigate evolutionary relationships, and examine interactions between RNA and proteins, ligands, or other nucleic acids. Structural information is also important in the study of riboswitches, which change their conformations in response to specific molecules, and ribozymes, which catalyze biochemical reactions. In medical research, RNA structural analysis can help investigate how sequence variants affect RNA stability, splicing, translation, or molecular interactions. In biotechnology, structural information supports RNA design, RNA-based therapeutics, and the development of molecules intended to bind particular RNA targets.
- Computational analysis of RNA structures involves several complementary methods. Secondary-structure prediction tools estimate which nucleotides are likely to pair, often using thermodynamic models, comparative sequence analysis, or machine-learning methods. Covariance models help detect conserved RNA families, while three-dimensional comparison tools examine structural similarity, nucleotide interactions, and recurring motifs. Molecular visualization software allows researchers to inspect RNA structures, display base pairs, examine ligands, and investigate interactions with proteins. Programmatic access and bulk downloads enable the integration of RNA structure information into automated bioinformatics pipelines. For reproducibility, researchers should record database identifiers, sequence versions, structure accessions, software versions, model confidence information, and the processing steps used in an analysis.
- Despite their value, RNA structure databases have limitations. Experimental structures may contain unresolved regions, incomplete chains, or only a subset of the full biological molecule. The conformation observed under experimental conditions may not represent every functional state of the RNA. Computationally predicted structures can be uncertain, particularly for long RNAs, flexible regions, pseudoknots, complex molecular assemblies, and RNA molecules that interact with proteins or ligands. Secondary-structure diagrams also simplify RNA architecture and do not fully represent three-dimensional interactions. Researchers should therefore examine experimental evidence, structural quality, sequence coverage, model provenance, and validation information before drawing biological conclusions.
- Choosing the appropriate database depends on the research objective. The PDB is a primary starting point for experimentally determined three-dimensional RNA structures, NAKB and RNA 3D Hub are useful for specialized nucleic acid and RNA structural annotations, RNAcentral supports integrated RNA sequence retrieval and secondary-structure exploration, and Rfam is particularly valuable for identifying conserved RNA families and consensus structures. NuFold can be considered when predicted three-dimensional RNA models are useful, while CoDNaS-RNA supports investigations of conformational diversity. Combining these resources provides a more complete view of RNA sequence, structure, evolution, and function than relying on any single database.
- In summary, RNA structure databases are fundamental resources in modern bioinformatics because they make structural information accessible for RNA discovery, classification, visualization, comparison, and functional interpretation. By integrating experimentally determined structures, predicted models, secondary-structure annotations, conserved RNA families, and structural motifs, these databases help researchers understand how RNA molecules function in biological systems. Their effective use requires careful attention to data provenance, structural quality, prediction uncertainty, database identifiers, and the distinction between experimentally observed and computationally inferred structures.