RNAdistance compares RNA secondary structures by computing distance metrics that quantify structural differences. Part of the ViennaRNA Package, it calculates tree edit distance and base pair distance between two structures represented in dot-bracket notation. These metrics help assess how similar or different two RNA conformations are, which is essential for studying structural evolution, evaluating prediction accuracy, and analyzing the effects of mutations on RNA folding.
ProteinIQ provides browser-based access to RNAdistance with no command-line setup required.
| Input | Description |
|---|---|
Structure 1 | First RNA secondary structure in dot-bracket notation (e.g., ((((....))))) |
Structure 2 | Second RNA secondary structure in dot-bracket notation (e.g., ((((.....))))) |
Structures must use standard dot-bracket notation where dots represent unpaired bases and matched parentheses represent base pairs.
| Column | Description |
|---|---|
Structure 1 | The first input structure |
Structure 2 | The second input structure |
Tree Edit Distance | Minimum cost to transform one structure's tree representation into the other |
Base Pair Distance | Number of base pairs present in one structure but not the other |
RNAdistance implements two complementary distance measures:
The simplest metric counts base pairs that exist in one structure but not the other. If structure A has the pair (i,j) and structure B does not, that contributes to the distance. The total equals the number of base pair insertions and deletions needed to convert one structure into the other.
For equal-length structures on identical sequences, base pair distance provides an intuitive measure: a distance of zero means identical folding, while higher values indicate greater structural divergence.
RNA secondary structures can be represented as ordered trees, where each node corresponds to a structural element. Tree edit distance computes the minimum-cost sequence of node insertions, deletions, and replacements needed to transform one tree into another.
The algorithm assigns costs to operations:
Tree editing captures structural relationships more faithfully than string-based metrics because it respects the hierarchical nesting of helices, loops, and bulges.
