Correctly predicting RNA 3D structures starting from sequence is a major, unsolved challenge in bioinformatics. An important sub-goal is the inference of the 3D structures of recurrent hairpin and internal RNA 3D motifs that appear as unpaired "loops" on secondary structure diagrams. RecurrentĀ 3D motif functions include:

  • Architectural roles introducing bends in helices (e.g. kink-turns) or changing helical twist (e.g. C-loops)
  • Anchoring RNA tertiary interactions (e.g., GNRA loops and loop-receptors)
  • Providing sites for proteins or small molecules to bind.

Analysis of 3D structures (for example using WebFR3D) shows that different RNA sequences can form the same RNA 3D motif. Since not all sequence variants of a given motif are present in the 3D database, accurate methods are needed to predict which sequences are likely to form known 3D motifs. This is the purpose of JAR3D.

Input and Output

JAR3D recognizes multiple types of input (see Examples above). If a secondary structure is provided, it is used to extract internal and hairpin loops. Unfolded single sequences are folded by UNAfold. Multiple sequence alignments without secondary structures are folded with RNAalifold. To specify chain break in internal loops, use the * symbol. Sequence(s) without asterisks and shorter than 25 nucleotides are interpreted as hairpins. Internal and hairpin loops should include closing Watson-Crick basepairs.

The output shows the top scoring motifs from the RNA 3D Motif Atlas. The user can view scoring statistics, visualize representative motif instances that potentially match input sequences.

Only internal loops are supported at this time. Please check back soon.


  1. We extract all hairpin and internal loops from a non-redundant set of RNA 3D structures from the PDB/NDB and cluster them in geometrically similar families.
  2. For each recurrent motif, we construct a probabilistic model for sequence variability based on a hybrid Stochastic Context-Free Grammar/Markov Random Field (SCFG/MRF) method we developed.
  3. To parameterize each model, we use all instances of the motif found in the non-redundant dataset and RNA knowledge of nucleotide interactions, especially isosteric basepairs and their substitution patterns.
  4. Given the sequence of a hairpin or internal loop from a secondary structure as input, each SCFG/MRF model calculates the probability that the sequence forms a given 3D motif. If the score is in the same range as sequences known to form the 3D structure, we infer that the new sequence can form the same 3D structure.
    Read more