RNAduplex

What is RNAduplex?

RNAduplex computes the hybridization structure between two RNA strands by predicting only intermolecular base pairs. Given two RNA sequences, it identifies optimal binding sites and calculates the hybridization free energy.

The program belongs to the ViennaRNA Package and uses the "interaction-only" (IO) approach: it models duplex formation without considering how intramolecular folding within each strand might compete with hybridization. This simplification makes RNAduplex extremely fast, operating at speeds comparable to sequence alignment while providing more information than simple BLAST searches. The tradeoff is that it may overestimate binding stability for sequences with stable internal structures.

When to use RNAduplex

RNAduplex works well for:

• Locating miRNA binding sites on mRNA targets
• Screening for potential RNA-RNA interactions across many sequence pairs
• Initial identification of antisense oligonucleotide targets
• Rapid pre-filtering before more detailed analysis with RNAcofold or RNAup

How to use RNAduplex online

Paste two RNA sequences, set a folding temperature, and RNAduplex returns the optimal intermolecular hybridization structure with binding site coordinates and free energy. Results appear as a spreadsheet row with dot-bracket notation showing exactly which nucleotides pair across the two strands.

Inputs

Supported file formats: .fasta, .fa, .txt

Settings

Output

Results display in spreadsheet format with the following columns:

Interpreting the structure notation

The output uses dot-bracket notation with an ampersand separating the strands:

.(((((.(((...((((((((((.&)))))))))))))))))).

• . indicates unpaired nucleotides
• ( and ) indicate paired bases (matching pairs face each other across the &)
• & marks the boundary between the two sequences

How RNAduplex works

RNAduplex uses dynamic programming to find the minimum free energy (MFE) configuration of intermolecular base pairs. The algorithm considers:

1. Watson-Crick and wobble pairs: Standard A-U, G-C, and G-U base pairing
2. Stacking energies: Stabilization from consecutive base pairs
3. Dangling ends: Energy contributions from unpaired nucleotides adjacent to helices
4. Temperature effects: Thermodynamic parameters rescaled to the specified temperature

The calculation deliberately excludes intramolecular structure formation. Each nucleotide can participate in at most one intermolecular base pair, but neither strand folds onto itself during the prediction.

Limitations

Because RNAduplex only considers intermolecular pairs, it does not account for:

• Competition from internal structure: If either RNA has stable intramolecular base pairs, the predicted binding site may be inaccessible
• Structure opening costs: The energy required to unfold existing structure before hybridization can occur

For more accurate predictions that include these effects, follow up RNAduplex hits with RNAcofold (full dimer folding including intramolecular pairs) or RNAup (adds accessibility costs to the calculation).