RNA secondary structure prediction

What is RNA secondary structure prediction?

RNA secondary structure prediction is the computational inference of base-pairing patterns within an RNA molecule, usually represented as stems, hairpins, bulges, internal loops, junctions, and unpaired regions. Thermodynamic methods estimate low-free-energy structures and ensembles, while comparative methods use conserved pairing and covariation across homologous sequences.

Use secondary-structure prediction when the scientific question concerns which nucleotides pair, whether a region is accessible, how a mutation changes a local fold, or which alternative structures are thermodynamically plausible. Minimum free energy is one summary of a structural ensemble, not proof that the RNA adopts one rigid conformation.

Sequence length, temperature, salt assumptions, pseudoknots, ligand binding, modifications, cotranscriptional folding, and cellular partners can change the observed structure. Add probing constraints or comparative evidence when available, and compare ensemble-level outputs instead of relying only on a single drawing.

When to use RNA secondary structure prediction

• Best fit. Base-pair maps, local accessibility, regulatory motifs, construct design, and mutation comparison
• Required context. RNA sequence plus relevant temperature, alignment, constraints, modifications, and experimental context

Benefits of RNA secondary structure prediction

• Fast sequence-level screening. Fast sequence-level screening while keeping the method and its native evidence explicit.
• Explicit base-pair and energy outputs. Explicit base-pair and energy outputs while keeping the method and its native evidence explicit.
• Supports ensemble and accessibility analysis. Supports ensemble and accessibility analysis while keeping the method and its native evidence explicit.

Primary limitations

• Model and condition dependence. Model and condition dependence; address it with appropriate context, controls, and orthogonal validation.
• Pseudoknots may be excluded. Pseudoknots may be excluded; address it with appropriate context, controls, and orthogonal validation.
• Cellular context is incomplete. Cellular context is incomplete; address it with appropriate context, controls, and orthogonal validation.

RNA secondary structure prediction methods

Nearest-neighbor thermodynamic models score loops and stacked base pairs, then use dynamic programming to find a minimum-free-energy structure or calculate a partition function over an allowed structure space. RNAsubopt samples or enumerates structures near the optimum, while RNAplfold focuses on local pairing and opening probabilities.

Comparative prediction asks whether homologous sequences preserve a common structure, including compensatory substitutions that maintain pairing. RNAalifold is appropriate for a real multiple alignment; it should not be substituted for single-sequence folding or used with unrelated sequences.

RNA secondary structure prediction applications

Common applications include evaluating untranslated regions, riboswitches, aptamers, guide RNAs, RNA therapeutics, splice-regulatory elements, primers, and mutations that may alter local structure or accessibility.

For long transcripts, local methods can identify stable windows and accessible regions without claiming a unique global fold. For interaction studies, secondary structure is often an input to accessibility-aware target prediction rather than the final answer.

How to run RNA secondary structure prediction online

Open the connected workflow, review the input roles and model assumptions, then keep every method’s native result separate during comparison.

1. Prepare sequence. Paste or upload the RNA sequence in FASTA format and confirm orientation, alphabet, length, and construct boundaries.
2. Set conditions. Set temperature and other supported folding options; add an alignment or experimental constraints only when the selected method accepts them.
3. Run predictors. Run RNAfold for the minimum-free-energy structure and ensemble metrics, then RNAsubopt or ViennaRNA for alternative structures when needed.
4. Compare ensembles. Compare dot-bracket structures, energies, pairing probabilities, ensemble diversity, and locally accessible regions rather than selecting by appearance.
5. Export evidence. Export structures, tables, plots, settings, sequence identifiers, and software versions for downstream review.

How to interpret RNA secondary structure prediction results

Report the sequence, temperature, parameter set, constraints, and whether energies refer to a minimum structure, an ensemble, or an interaction. A small energy difference between structures can indicate competing conformations rather than a decisive winner.

Pairing probabilities and ensemble diversity provide uncertainty information that a single dot-bracket string omits. Validate important motifs with covariation, chemical probing, mutational evidence, or structural experiments whenever the decision is consequential.

How RNA secondary structure prediction works

Run RNAfold, RNAsubopt, and the curated ViennaRNA interface in parallel so minimum-energy, ensemble, and alternative-fold evidence remain distinguishable.

1. Prepare sequence. Paste or upload the RNA sequence in FASTA format and confirm orientation, alphabet, length, and construct boundaries.
2. Set conditions. Set temperature and other supported folding options; add an alignment or experimental constraints only when the selected method accepts them.
3. Run predictors. Run RNAfold for the minimum-free-energy structure and ensemble metrics, then RNAsubopt or ViennaRNA for alternative structures when needed.
4. Compare ensembles. Compare dot-bracket structures, energies, pairing probabilities, ensemble diversity, and locally accessible regions rather than selecting by appearance.
5. Export evidence. Export structures, tables, plots, settings, sequence identifiers, and software versions for downstream review.

Inputs and outputs

Check formats before running, then inspect and download the result from every workflow step.

Inputs

• RNA inputs. FASTA TXT One or more RNA sequences in FASTA, with optional alignment or method-supported constraints.

Outputs

• Prediction evidence. CSV JSON TXT SVG PDB mmCIF ZIP Dot-bracket structures, free energies, ensemble metrics, tables, plots, and downloadable native files.